Viewpoints
A roadmap for research on
crassulacean acid metabolism
(CAM) to enhance sustainable
food and bioenergy production in
a hotter, drier world
Summary
Crassulacean acid metabolism (CAM) is a specialized mode of
photosynthesis that features nocturnal CO2 uptake, facilitates
increased water-use efficiency (WUE), and enables CAM plants to
inhabit water-limited environments such as semi-arid deserts or
seasonally dry forests.Humanpopulationgrowthandglobal climate
change now present challenges for agricultural production systems
to increase food, feed, forage, fiber, and fuel production. One
approach to meet these challenges is to increase reliance on CAM
crops, such asAgave andOpuntia, for biomass production on semi-
arid, abandoned, marginal, or degraded agricultural lands. Major
research efforts are now underway to assess the productivity of
CAM crop species and to harness the WUE of CAM by engineering
this pathway into existing food, feed, and bioenergy crops. An
improved understanding of CAM has potential for high returns on
research investment. To exploit the potential of CAM crops and
CAMbioengineering, it will be necessary to elucidate the evolution,
genomic features, and regulatory mechanisms of CAM. Field trials
and predictive models will be required to assess the productivity of
CAM crops, while new synthetic biology approaches need to be
developed for CAM engineering. Infrastructure will be needed for
CAM model systems, field trials, mutant collections, and data
management.
I. Introduction
Two of the grand challenges facing our society in the twenty-first
century are: the continuing rapid expansion of the world’s human
population, now at 7.2 billion, which is expected to increase by
33‒71% by 2100 (Gerland et al., 2014); and the potential increase
in the frequency and intensity of drought, along with decreases in
soil moisture, related to global climate change (Dai, 2013; Cook
et al., 2014) (Fig. 1). These two externalities could seriously impact
future food and energy security while increased competition for
land and water resources between urban growth and agricultural
production systems will intensify demands for limited freshwater
resources. Fortunately, a viable solution to these challenges exists in
crassulacean acid metabolism (CAM), a specialized type of
photosynthesis that results in enhanced plant water-use efficiency
(WUE). With an inverted day : night pattern of stomatal closure/
opening relative to the more typical C3 and C4 crops, the WUE of
CAMplants can be six-fold higher than that of C3 plants and three-
fold higher than that of C4 plants under comparable conditions
(Borland et al., 2009). Most present-day food crops (e.g. rice
(Oryza sativa L.), corn (Zea mays L.)) and bioenergy crops (e.g.
poplar (Populus spp.), switchgrass (Panicum virgatum L.), sugar-
cane (Saccharum spp.)) useC3 orC4 photosynthesis, whereas CAM
crops have yet to be extensively adopted and developed.
Two strategies could be used to explore the potential of CAM
for food and biomass production: the development of CAM
crops as new sources for food and biomass; and the transfer of
CAM machinery into existing food and biomass crops (Fig. 1).
Multiple CAM species are currently used as food sources that
provide fruits, vegetables, and various natural products (Support-
ing Information Table S1). Some CAM plants (e.g. Agave spp.)
have potential as biofuel crops due to their high theoretical
biomass yield (Davis et al., 2014) and low recalcitrance for biofuels
conversion (Li et al., 2014). The use of CAM species or the
application of engineered CAM to improve plant WUE could
curtail crop losses under catastrophic episodes of heat and drought
and contribute to the expansion of crop production into
abandoned or semi-arid lands. Here we outline a research
roadmap that identifies some important scientific questions in
CAM research, and provides direction for realizing the potential of
CAM for human good in terms of food, feed, fiber, and fuel
production. The infrastructure needs for further developing the
CAM research community are discussed.
II. Research questions
1. How did CAM evolve from a C3 ancestor?
CAM has evolved multiple times in diverse lineages of vascular
plants and is found in over 400 distinct genera across 36 families
(J. A.C. Smith et al., unpublished).However, our understanding of
the evolutionary history of CAM is still rudimentary. There are
several reasons for this, all of which present formidable, yet
surmountable, challenges.
First, there is continuing debate about how exactly to define a
CAM plant (Winter et al., 2015). Numerous surveys of succulent
plants have provided evidence of a clear bimodal distribution of
13C : 12C isotope ratios, with a minimum in the frequency
distribution typically observed at a d13C value of c. 
20&.
Species with d13C values less negative than 
20& correspond to
obligate CAM plants engaged in fixing the majority of their CO2

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504 491
www.newphytologist.com
Forum
at night. However, d13C values between 
20& and 
27&
might indicate a small but significant degree of nocturnal CO2
fixation (Winter & Holtum, 2002). The ability to carry out some
dark CO2 fixation while still primarily engaged in C3 photosyn-
thesis is very likely a key intermediate step along the C3-to-CAM
evolutionary trajectory. However, very little is known about the
prevalence and phylogenetic distribution of this low-level CAM
activity, which can only be detected by direct measurements of
CO2 exchange and acidity fluctuations on living accessions under
conditions conducive to the expression of CAM.
Second, many CAM-evolving groups are also spectacularly
diverse. Scoring the presence or absence of CAM has been
accomplished for over 1000 species of orchids (Silvera et al., 2009,
2010), yet this covers only a fraction of all orchid species
(> 25 000). A study of nearly 2000 species of bromeliads represents
the single largest carbon-isotope survey to date (Crayn et al., 2004,
2015), corresponding to almost two-thirds of the family; by
contrast, phylogenies of the Bromeliaceae have so far contained
fewer than 200 taxa, including both C3 and CAM species
(Givnish et al., 2011, 2014; Silvestro et al., 2014). Other CAM
groups are equally daunting: a recent attempt to reconstruct CAM
evolution in the genus Euphorbia (c. 2000 species) has provided
valuable insight into a previously understudied family (Horn
et al., 2014), but only c. 10% of the genus was sampled. On the
other hand, the suborder Portulacineae of Caryophyllales
(Arakaki et al., 2011; Edwards & Ogburn, 2012), which currently
has a relatively more complete phylogenetic sampling, lacks an
equivalently fine-grained survey of CAM capability. To under-
stand the evolutionary dynamics of CAM origins and losses, a
complete sampling of both phylogeny and phenotype is required.
Thus, there is a strong need for the strategic development of
specific lineages as model systems (see Section III.1) coupled with
genomics research (see Section II.2) to infer the evolutionary
trajectory of CAM.
(a)
(c)
(d)
(b)
(e)
Fig. 1 Challenges and crassulacean acid metabolism (CAM) solution. (a) Expected geographical population changes in millions from 2013 to 2100 (A. E.
Raftery, University ofWashington, 2013, based onmethodology described in Raftery et al. (2012, 2013, 2014)). (b) Percentage changes from 1980–1999 to
2080–2099 in themulti-model ensemblemean soil-moisturecontent in the top10 cm layer (Dai, 2013). (c) TheproductionofC3 crops is negatively impactedby
drought stress, rice image © 2013 Techin24. (d) Water-use efficient CAM crop (e.g. Agave) for biomass production on marginal land, Agave image © 2014
jferrer; and (e) CAM plants demandmuch less water than C3 or C4 plants (Borland et al., 2009); thus, CAM crops and non-CAM crops engineered with CAM
provideanexcellentwater-saving strategy to address thegrand challenges causedby future increases in bothworldpopulationanddrought stress, rice image©
2012 KhartesevaTetiana, maize image © 2011 tilo, cactus image © 2008 andylin.
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist492
2. What are the genomic features of CAM plants?
The growth and development of CAM plants are controlled by
functional elements encoded in the genome. Identification of the
genomic features of CAM plants will benefit greatly from the
construction of the pan-genomes (Hirsch et al., 2014) of CAM
species from diverse lineages that encompass obligate CAM species
and species with a facultative (inducible) component of CAM.
Comparative genomics approaches will allow the identification of
both the core genes that are shared by different CAM species and
the genes that are specific to different biochemical CAM types
(Holtum et al., 2005). Three important types of functional
elements are embedded in genomic sequences: protein-coding
genes, noncoding RNA (ncRNA) genes, and regulatory
cis-elements. Protein-coding genes express mRNAs that encode
translational information for protein synthesis. The ncRNA genes,
which are not translated into proteins, produce transcripts that
function directly as structural, catalytic, or regulatory RNAs (Eddy,
2001). The cis-elements play important roles in regulating the
expression of protein-coding genes and ncRNA genes. The
neofunctionalization and subfunctionalization of at least some of
the genes required for CAM are likely to have occurred through
the differentiation of cis-regulatory elements that control the
magnitude and patterns of gene expression (Monson, 2003;
Hibberd & Covshoff, 2010). Equally important are mutations or
polymorphisms within the protein-coding regions that result in
modified functional domains that might have been necessary to
adjust the kinetic properties of enzymes and transporters required
for CAM.
Ongoing and future genome projects should identify genes
recruited specifically to CAM function and their associated
cis-elements through analysis of conserved noncoding sequences
among co-expressed genes within species, or orthologous genes
shared between different CAM species. The cis-regulatory elements
identified via computational analysis should be validated using
reporter genes, such as GUS (for tissue-specific expression), GFP
(for cell-specific expression), and LUC (for temporal expression).
Special attention should be given to promoters responsible for
drought-inducible CAM expression in facultative CAM plants
such asClusia pratensis Seem. andMesembryanthemum crystallinum
L., in addition to those controlling temporal and cell-specific gene
expression. The comparison of diverse CAM genomes should
explain the degree of evolutionary flexibility that allowed this
complex trait to emerge. For example, have the same gene orthologs
been recruited to a CAM function in different species, or have
different orthologs been recruited (Christin et al., 2015)? Applying
a commonly agreed-upon set of criteria for identifying such
genomic features will be critical for answering this question. A
comparative framework for identifying genomic elements relevant
to CAM is illustrated in Fig. 2.
Agave tequilana (oCAM)
Kalanchoë laxiflora (oCAM) 
Pineapple (oCAM)
Clusia pratensis (fCAM)
Ice plant (fCAM)
Arabidopsis thaliana (C3)
Populus trichocarpa (C3)
Rice (C3)
Homolog groups in the CAM pan-genome
Y
Z
Co
m
m
on
C 3 Co
re
 C
AM
In
du
ci
bl
e 
CA
M
Ob
lig
at
e 
CA
M
•   Phylogeny of genes in the homolog group
•   Gene expression (RNA, protein)
•   DNA polymorphism
•   Cis-regulatory elements
•   3D protein structure
X
Y
Y
Y
Y
Y
Y
Y
Y
— Y
Y
Y
Y
Y
—
—
—
—
—
—
Y
Y
—
—
—
Y
Y
Y
—
—
—
—
—
Y
—
—
—
—
—
—
—
—
Y
—
—
—
—
—
—
—
—
Y
—
—
—
—
—
—
—
—
Y
—
—
—
—
—
—
—
—
Y
—
—
—
—
—
—
—
Y
Y
Y
Fig. 2 Genomic elements of crassulacean acidmetabolism (CAM) plants in a comparative framework. TheX-axis represents the pan-genome,which is divided
into homolog groups including the groups (i.e. ortholog groups) shared by two or more species/genomes and the groups (i.e. paralogs) representing species-
specific gene expansion families. The Y-axis represents the plant species, which can be categorized as facultative CAM (fCAM), obligate CAM (oCAM), or C3
photosynthesis (C3).C3 species areusedas comparators to identifyCAM-relatedgenes.TheZ-axis represents the important information relevant to thegenes in
each individual homolog group (e.g. gene expression, DNA polymorphism, post-translational modification, cis-regulatory elements, epigenetic modification,
three-dimensional (3D) protein structure). Y, the homolog group contains gene(s) from the corresponding species;―, the homolog group does not contain
gene(s) from the corresponding species. Thehomologgroups canbedivided into six categories: (1) common, homologgroups sharedbyC3, fCAM, andoCAM
species; (2) C3-specific, homolog groups shared only amongC3 species; (3) core CAM, homolog groups shared only by fCAMand oCAM species; (4) inducible
CAM, homolog groups shared only by fCAM species; (5) obligate CAM, homolog groups shared only by oCAM species; and (6) species-specific, species-
specific gene expansion families.

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 493
3. What are the molecular mechanisms regulating CAM?
CAM regulation has been studied from numerous perspectives
including light–dark and circadian clock control over each 24-h
cycle, as well as the developmental and abiotic stress-dependent
regulation of the establishment of CAM (Cushman & Bohnert,
1999; Hartwell, 2006; Freschi & Mercier, 2012). CAM requires
strict temporal control of the associated metabolism in order to
prevent futile cycling between dark-period CO2 fixation to malate
and light-period malate decarboxylation. Furthermore, the signal
transduction pathways that control inverse stomatal opening and
closing in CAM plants must be deciphered in detail, as there is
currently very little direct experimental data relating to stomatal
guard cell signaling in CAM species. Also, redox control could be
critical for synchronization of metabolism and transport over the
diel CAM cycle, as suggested by the observation that some enzymes
involved in theCalvin–Benson cycle inC3 andC4plants, algae, and
cyanobacteria are regulated by thioredoxins to achieve higher
activities in reduced states than in oxidized states (Michelet et al.,
2013).
In amaturemaize leaf, a C3-to-C4 developmental gradient exists
between the leaf base (C3) and the leaf tip (C4) (Li et al., 2010). In
Agave americana var. marginata Trel., young and mature leaves on
the same plant use the C3 and CAM photosynthetic pathways,
respectively (X. Yang et al., unpublished data). It would therefore
be very useful to compare the gene expression patterns between
C3 and CAM leaf tissue within the same CAM plant in order
to understand the developmental regulation of CAM. To
understand abiotic stress-dependent regulation of CAM, studies
must be undertaken using truly facultative CAM species (e.g.
M. crystallinum,C. pratensis) to identify the regulatory components
required for the drought induction of CAM, and the subsequent
return to C3 under well-watered conditions.
Achieving a comprehensive understanding of CAM-associated
regulatory mechanisms will benefit greatly from the application of
functional genomics approaches that include both computational
predictions based on omics data and experimental characterization
using molecular and genetics tools (Fig. 3). Dissecting the
complexity of CAM regulation will be facilitated by the construc-
tion of protein–protein interaction and gene regulatory networks
(GRNs), which are the collection of interactions between
transcription factors and their target genes. The integration of
complete genome sequences with RNA-seq and chromatin
immunoprecipitation sequencing (ChIP-seq) experiments offers
an exciting platform to dissect and model GRNs using computa-
tional approaches such as Bayesian inference, Boolean modeling,
linear and nonlinear regression methods, Granger causality-based
inference, and cross-correlation analysis (Wallach et al., 2010;
Marbach et al., 2012; Middleton et al., 2012; Krouk et al., 2013;
Moghaddam & Van den Ende, 2013; Tam et al., 2013).
Despite recent and continuing advances with a range of omics
projects ongoing in CAM species (Borland et al., 2014), a key area
that remains lacking is the study of post-translationalmodifications
associated with CAM regulation. Reversible phosphorylation of
phosphoenolpyruvate carboxylase (PPC) by its specific, circadian
clock-controlled protein kinase, PPCK, is one of the few CAM
regulatory steps understood in any detail (Hartwell et al., 1999;
Taybi et al., 2000; Boxall et al., 2005; Dever et al., 2015). In the
coming years, research programs focused on achieving a compre-
hensive understanding of CAM regulationmust determine the role
of regulatory processes such as post-transcriptional modification of
mRNA stability or translatability and post-translational modula-
tion of protein activity (e.g. phosphorylation/dephosphorylation,
ubiquitination, glycosylation).
4. How might CAM be engineered into C3 or C4 plants?
As discussed in Section I, CAM-engineering is a viable strategy to
improve WUE in existing non-CAM crops for food and biomass
production in dryland areas. In principle, CAM-into-C3/C4
engineering is realistic because: (1) CAM has evolved from diverse
C3 species via convergent or parallel evolution (see Section II.1); (2)
the existence of facultative CAM species, in which CAM can be
induced from C3 or C4 by drought or salt stress (see Section III.1),
suggests that no incompatibilities exist betweenCAMandC3/C4 at
the organismal level; and (3) CAM is a single-cell carbon
concentrating mechanism that does not require differentiated
mesophyll and bundle sheath cell types, each with their own
specialized metabolic adaptations. Ideally the C3 target species for
CAM engineering should meet the following criteria: (1) a genome
that has been fully sequenced and well annotated; (2) an easily
transformed species with a well-established stable transformation
protocol; (3) a large impact on food or bioenergy production; and
(4) a crop that is currently not well suited for production on
dryland. Poplar and rice are examples of such candidate target C3
crops for CAM engineering, representing bioenergy and food
crops, respectively. If the CAM-into-C3 engineering effort is
successful, the potential of CAM-into-C4 engineering can be
investigated as a means to further enhance the WUE of major C4
crops such as corn and sorghum (Sorghum bicolor (L.) Moench).
Engineering of CAM into C3 crops will require a temporal
reprogramming (e.g. diel-cycle shift) of the expression of genes
shared between the C3 and CAM pathways, transferring CAM-
specific genes, possibly modifying endogenous genes (i.e. silencing
or knockout) in the host, engineering of leaf anatomical traits (e.g.
succulence, cell size, intercellular air space), and most likely an
inducible system that would initiate CAMwhen desired (e.g. under
drought stress). Currently, the exact number of genes needed to
introduce CAM into a C3 species remains unclear; however,
multiple CAM-related genes will need to be manipulated in a
modular manner, including: (1) a carboxylation module for CO2
fixation and nocturnal accumulation of malic acid in the vacuole;
(2) a decarboxylation module for release of CO2 frommalate; (3) a
stomatal control module for nocturnal stomatal opening and
stomatal closure during the daytime; and (4) an anatomicalmodule
for increasing leaf succulence (Borland et al., 2014). Furthermore,
these fourCAMmodules need to be integrated to establish CAMas
an efficient system inC3plants.Hypothesizedminimal gene sets for
the carboxylation and decarboxylation CAM modules are listed in
Table S2. Elucidation of the equivalent gene lists for stomatal
control and succulence must await detailed studies of these
processes in CAM species, as the required data are currently
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist494
(a)
(b)
(c)
Fig. 3 An integrative functional genomicsapproach for crassulaceanacidmetabolism(CAM)plants. (a)Omicsdata (i.e. genomics, transcriptomics, proteomics,
metabolomics) are generated for CAMplants, funnel image© 2013 TimArbaev,Opuntia image© andylin,Agave image© SSSCCC, pineapple image© 2012
julichka. (b) The omics data are analyzed to predict CAM-related genes and putative gene (regulatory) networks; and (c) various experimental approaches are
used to characterize the CAMgenes predicted by approaches in (b). [CH2O]n, carbohydrates; OAA, oxaloacetate; PEP, phosphoenolpyruvate; RuBP, ribulose
1,5-bisphosphate; triose-P, triose phosphate; TF, transcription factor; Y1H, yeast one-hybrid; Y2H, yeast two-hybrid; BiFC, bimolecular fluorescence
complementation.

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 495
lacking. Potential crosstalk between the CAMmodules needs to be
considered. For example, a reduction in the partial pressure of CO2
inside the leaf (pi) due to PPC activity in the dark following the
successful introduction of a functioning carboxylation module
could result in nocturnal stomatal opening whereas an increase in pi
due to CO2 released from malate by the decarboxylation module
during the daytime could induce stomatal closure. Thus, insertion
of additional genes for a stomatal controlmodulemight be obviated
by the successful installation of diel malate turnover in the leaf
groundmesophyll. CAMengineering is well beyond the capacity of
traditional plant biotechnology that is limited to transferring and
controlling only a few genes. Synthetic biology offers the potential
to address the challenge of CAM engineering via new concepts and
toolboxes (DePaoli et al., 2014). The application of synthetic
biology to CAM engineering involves five steps: (1) establishment
of a parts library (e.g. genes, promoters, terminators, genetic
insulators); (2) circuit design; (3) assembly of multi-gene con-
structs; (4) transfer (i.e. in planta gene stacking); and (5) evaluation
of engineered plants (Fig. 4).Multiple iterations of steps 2–5will be
required to achieve optimized performance of engineered CAM.
The parts information needs to be derived from knowledge of the
core CAM genes and regulatory mechanisms (see Sections II.2–
II.3) and informed by phylogenetic analyses (see Section II.1) that
identify independently evolving modules of traits. To prevent
influence by inappropriate or competing signals emanating from
their surrounding genomic environment, it is necessary to protect
transgenes with genetic insulators, which are a class of DNA
sequence elements with the ability to block the action of a distal
enhancer on a promoter or act as barriers to prevent the advance of
nearby condensed chromatin that might otherwise silence expres-
sion (West et al., 2002; She et al., 2010). Finding suitable insulators
for target C3 species is an important task for CAM-into-C3
engineering.
To streamline the downstream processes and facilitate collab-
oration in the CAM research community, a standard for the
construction of the gene parts and circuits should be established.
Circuit design can adopt a modular approach wherein the parts are
first assembled into carboxylation, decarboxylation, stomatal
control, and anatomical modules; these modules are then
connected into a CAM system. Various methods have been
developed for assembling multi-gene constructs (DePaoli et al.,
2014); however, none of them allow flexible, clean, and efficient
assembly of parts from a single universal library. New high-
throughputmethods for in vitro assembly ofmulti-gene constructs,
such as those described recently for mammalian systems (Guye
et al., 2013;Torella et al., 2014), are needed.A significant challenge
for transferring assembled CAM gene modules will be to insert
these large multi-gene constructs into the plant genome while
maintaining the structural and functional stability of the modules.
Methodologies that are site-specific, functional formultiple rounds
of targeted in vivo insertions, and compatible with multiple
methods of plant transformation still need to be developed.
Progress on both in vitro assembly of DNA parts and in planta gene
stacking for iterative insertion of marker-free DNA modules is
underway and merits further development (H. C. DePaoli et al.,
unpublished). The transgenic plants generated during CAM
engineering should be evaluated using omics approaches (e.g.
transcriptomics, proteomics, metabolomics, phenomics). These
Fig. 4 Crassulacean acid metabolism (CAM)
engineering using synthetic biology
approaches.
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist496
omics data can be used for system dynamics modeling (Borland &
Yang, 2013;Owen&Griffiths, 2013) and diel flux balance analysis
(Cheung et al., 2014) to inform metabolic and regulatory refine-
ments that will improve the performance of engineered CAM.
Moreover, the diel flux balancemodel could aid the optimal design
of the carboxylation, decarboxylation, and stomatal control
modules before they are engineered into a C3 species.
5. How can sustainable CAM crop production systems be
established?
CAM crops such as Agave, Opuntia, and pineapple (Ananas
comosus (L.) Merr.) have potential as biofuel and food crops on
abandoned, marginal, and degraded land in light of published
reports on their high productivities (Table S1). Agave and Opuntia
species have been used traditionally in a wide range of foods,
beverages, food products, forage, fodder, and also dietary
supplements, pharmaceuticals, and cosmetics. The many indus-
trial uses of Agave fibers include cordage, textiles, construction
materials, and solid fuels (Cushman et al., 2015). However, more
extensive field trials are required to provide data for the food, feed,
fiber, and bioenergy uses of CAM species to integrate into a
framework that considers sustainable yields given externalities of
land availability, management inputs, economics, and market
demand (Davis et al., 2011; Nunez et al., 2011; Yan et al., 2011;
Lewis et al., 2015). Davis et al. (2011) estimated that substantial
abandoned agricultural land exists globally and suggested that this
land could be reclaimed and repurposed for bioenergy production.
Furthermore, in areas where CAM is a viable option, biomass
production should be evaluated against sustainability metrics that
include water quality, water use, fertilizer inputs, potential
herbicide and pesticide applications, and biodiversity (McBride
et al., 2011). Such information would be useful for resource
assessment models and for evaluating environmental consequences
of CAM plantations.
System-level analysis of agricultural production has been applied
to many agricultural production settings and should be similarly
developed for CAM plants (Davis et al., 2015). To date, there has
been only one detailed life-cycle analysis (LCA) that addresses a
CAM crop (Yan et al., 2011). That study concludes that an Agave-
based bioenergy system would have greater energy returns per unit
of energy input than a bioenergy system based onmaize.Moreover,
greenhouse gas emissions per unit of energy produced from Agave
would be much lower than those from a maize grain system (Yan
et al., 2011). Physiological models of CAM also require the
development of tools comparable to those used for C3 andC4 crops
(Davis et al., 2015).To assess the relative benefits ofCAMcropping
systems, studies should undertake comparative physiological
modeling and LCA for obligate CAM crops and other bioenergy
crops including Jatropha (C3), poplar (C3), willow (Salix spp.; C3),
Miscanthus (C4), sugarcane (C4), and switchgrass (C4).
In addition, productivity models could be valuable tools for
identifying management scenarios suitable for sustainable crop
production. Suchmodels exist for only a few bioenergy crops (Nair
et al., 2012) and have proven useful in efforts to tailor crops and
cropping systems to various environments (Miguez et al., 2012).
The most widely used model for CAM plants is the environmental
productivity index (EPI), which estimates potential yield based on
temperature, soil water, and solar radiation (Nobel, 1984; Garcia
de Cort
azar & Nobel, 1990). Owen & Griffiths (2014) have
developed a geospatial model based on the EPI approach to predict
bioethanol yield potential for Agave and Opuntia species in
Australia, and used that model to predict crop production on low-
grade and marginal lands under current and future climate
conditions (Figs 5, S1). Simulations highlight that the same
WUE features of the CAM pathway which distinguish it from C3
and C4 bioenergy candidates also offer resilience to predicted
climate change.
Although the EPI has proven useful, it lacks mechanistic
details. Owen & Griffiths (2013) have thus developed a systems
dynamics model of CAM that integrates biochemical and
physiological constraints to predict leaf-level gas exchange and
titratable acidity fluctuations. While this model does not allow
physiological predictions at the canopy scale, key regulatory
components of this model could be manipulated to simulate
CAM expression across contrasting succulent life forms. The
model was able to identify parameters that limit carbon uptake
over the diel cycle and thus may prove useful as a tool to help
target synthetic biology approaches to improve crop production
(see Section II.4). Opportunities exist to incorporate gene
regulatory and metabolic networks into this model and to link
CAM expression to whole-plant traits such as net assimilation
rate, relative growth rate, and the allocation and partitioning of
carbon among plant components.
III. Infrastructure for the CAM community
1. Model systems for CAM research
Model systems are key elements of integrative research programs.
The development of three types of model systems is suggested:
phylogenetic lineages that include both C3 and CAM (and
potentially also C4) species for studying CAM evolution; CAM
species with a small genome, short-life cycle, and well-established
genetic transformation system for functional genomics research;
and CAM species with potential for food, feed, and biomass
production as model crops.
Potential model lineages with species showing C3, CAM and
C4 Ideally, model lineages would include species from multiple
CAM origins and variations in the operation of CAM. The
neotropical genus Clusia is the only genus of woody eudicotyled-
onous trees reported to use CAM (L€uttge, 2006, 2008). This genus
includes obligate CAM and C3 species as well as species that show
facultative CAM and reversible shifts between C3 and CAM
(Winter & Holtum, 2014). A unique opportunity exists with
Portulaca, a lineage that includes the only known examples of C4
species in which CAM can be induced by drought stress (Koch &
Kennedy, 1980, 1982; Christin et al., 2014). Understanding the
molecular, anatomical, and metabolic mechanisms that allow for
the co-existence of C4 and CAM could facilitate engineering of
both pathways into a single plant. Another good model lineage is

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 497
Erycina, a group of orchids containingmembers that performC3 or
CAM photosynthesis (Silvera et al., 2010).
Model species for functional genomics Two Kalancho€e species,
K. laxiflora Baker and K. fedtschenkoi R.-Hamet & Perrier, have
been established as obligate CAM model systems due to their
relatively small genome sizes, short life cycle, and amenability to
stable transformation (Aida & Shibata, 1996; Garces et al., 2007;
Garcia-Sogo et al., 2010; Dever et al., 2015). Their genomes are
currently being sequenced, assembled, and annotated (Table S3).
In addition, the common ice plant (M. crystallinum) is a well-
studied, facultative CAMmodel, in which CAM can be induced by
salinity or water-deficit stress (Winter & Holtum, 2005, 2007,
2014). M. crystallinum has played a seminal role in our under-
standing of the function and subcellular localization of enzymes
involved inCAMfunction (Holtum&Winter, 1982;Winter et al.,
1982), aswell as definingmany of the corresponding genes for these
enzymes (Cushman et al., 2008b) and intracellular transporters
(Kore-eda et al., 2013). The ice plant transcriptome and genome
are currently being sequenced, assembled, and annotated (Table
S3). Another emerging model is Sedum telephium L. (=Hylote-
lephium telephium subsp. telephium L.), a C3–CAM intermediate
(Groenhof et al., 1990; Borland, 1996). Genomic resources for the
genus Sedum have been developed (Chao et al., 2010; Gao et al.,
2013), and sequencing of the genome of S. telephium is in progress
(Table S3).
Model CAM crops Currently, only a limited number of CAM
crops (e.g. Agave spp., Manfreda spp., Polianthes spp.,
Prochnyanthes mexicana (Zucc.) Rose, Aloe vera (L.) Burm.f.,
A. comosus, Hylocereus spp., Opuntia ficus-indica (L.) Mill.,
Stenocereus spp., Vanilla planifolia Jacks. ex Andrews) have been
used for the production of food, bioenergy, fiber, and animal feed
(Table S1). The production scale of existing CAM crop is currently
much smaller than that of major C3 or C4 crops, although the
benefits of these CAM plants as cash crops are of considerable
importance to the economies of many nations in the tropics and
subtropics. Due to the potential increase in the frequency and
(a) (b)
Fig. 5 Predicted productivity of Agave tequilana and Opuntia ficus-indica under current and future climate conditions. (a) Simulations under current climate
conditions show that the geographical distribution of highly productive areas (environmental productivity index (EPI) > 0.5) is more restricted forA. tequilana
than forO. ficus-indicabecauseA. tequilanahasa comparativelyhigher sensitivity tonocturnal temperatureand lower capacity tobuffer against periodsof low
soil water potential. The higher saturation point for carbon uptake response to photosynthetically active radiation (PAR) of O. ficus-indica compared to
A. tequilana (35 vs 29mol m
2 d
1, respectively) has a negative impact on yields at latitudes > 30°S or 30°N. The productivity distribution of both species is
restricted to areas where tmin > 0°C; simulations used environmental inputs averaged over the period 1950–2000; EPI was scaled with a value for maximum
productivity (Pm) that could occur under irrigation and optimal planting density (44 and 46 Mg (dry) ha
1 yr
1 for A. tequilana and O. ficus-indica,
respectively). Productivity simulationsmay be linearly re-scaled for different values of Pm. (b) Simulated productivity under future climate conditions show the
percentage change in productivity between the present and worst-case climate scenario in the year 2070 (AR5 representative concentration pathway
8.5Wm
2, 70RCP8.5). Inside the range of latitudes from 30°S to 30°N, A. tequilana simulations suggest that climate change will have a greater negative
impact on productivity compared to O. ficus-indica. Outside this range, climate change has a beneficial impact on A. tequilana productivity. In general,
O. ficus-indica displays stronger resilience to climate impacts.
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist498
intensity of drought (Fig. 1), CAM crops could play an increasing
role inmeeting our future needs for food and bioenergy. Therefore,
we propose more widespread planting of CAM crops, with an
initial focus on three major CAM crops as models: Agave,Opuntia,
and pineapple. Agave spp. are economically important CAM crop
species, holding great potential for production of biofuel, fiber,
food, and animal feed in water-limited areas (Li et al., 2014; Nava-
Cruz et al., 2014). Agave was recently added to the list of potential
dedicated biomass crops in the United States (US DOE, 2014).
Given the importance ofA. tequilana F.A.C.Weber for commercial
alcohol production and as a representative for all agaves, its genome
is currently being sequenced, assembled, and annotated (Table S3).
Opuntia spp. (e.g.O. ficus-indica) have been introduced as forage
and fodder crops inmany semi-arid regions of the world (Russell &
Felker, 1987; Nobel, 1994; Le Hou
erou, 1996). The young
cladodes and fruits are not only consumed as food directly or as
diverse processed food items, but are also used in a wide range of
other products such as sweeteners, food coloring, dietary supple-
ments, cosmetics, and medicines (Feugang et al., 2006; Moßham-
mer et al., 2006). Large-scale transcriptome and genome
sequencing of this CAM species is in progress (Table S3).
Pineapple, the third most important tropical fruit after banana
and citrus, is cultivated in over 80 countries in tropical and
subtropical regions worldwide. In vitro plantlets of pineapple
perform C3 photosynthesis, while adult plants perform CAM
photosynthesis constitutively (Freschi et al., 2010). Both its
genome and transcriptome have been sequenced (Table S3).
While the development of sustainable production systems is
focused on the earlier three model CAM crops, the potential of
other CAM crops for food production should be exploited. In
addition to the crops listed inTable S1, it is necessary to evaluate the
food quality and yield of other CAM species in order to develop
new crops for the sustainable production of food and other high-
value products.
2. Field trials
Establishment of replicated field trials is critical for quantifying
yields under contrasting environmental conditions. Such efforts
would provide data essential to the development of empirical
models that use environmental conditions as inputs and give
probabilistic estimates of yield (unlike EPI) for CAM species. Such
models, based on field observations, could be used to: (1) validate
existing EPI-based models; (2) identify locations where sustainable
and profitable yields are possible; and (3) more accurately simulate
yield in a hotter, drier world under projected Intergovernmental
Panel on Climate Change (IPCC) climate scenarios. Field trials for
Agave are being conducted in Australia (Holtum & Chambers,
2010) and the United States (Davis, 2013), and should be
replicated in Mexico with local species, where > 200 varieties have
been developed for either fiber or alcohol production (Colunga-
Garc
ıaMar
ın & Zizumbo-Villarreal, 2007). A replicated field
study by the Nevada Agricultural Experiment Station is currently
underway to evaluate the productivity of three Opuntia species.
Beyond field trials, a network of common gardens for eachmajor
CAM crop should be established, with collections composed of
multiple genotypes from natural populations and planted in clonal
replicates in 3–4 alternate environmental conditions (e.g. soil
extremes, temperature minima and maxima, water availability).
Omics data (e.g. genomics, transcriptomics, metabolomics, phe-
nomics) can be collected for individual plants in the common
garden, which will serve as a foundation for unraveling the
association between genomic elements and trait phenotypes. Results
from such studies would be very useful to inform application of
genomic selectionmethods for genetic improvement inCAMcrops.
3. Genetic mutant collections
Mutant collections should be created for model CAM species to
facilitate functional genomics research noted earlier. Such a
collection has been created for M. crystallinum using fast-neutron
bombardment, which generates small deletions or rearrangements
within the genome, and has been used for the isolation and
characterization of CAM-deficient mutants (Cushman et al.,
2008a). Targeted loss-of-function Kalancho€e mutants generated
by RNAi-mediated gene silencing have been created for a wide
selection of candidate genes with functions in CAM (Dever et al.,
2015). Whole-genome sequencing of the genotypes within a
network of common gardens (see Section III.2) can also help to
discover naturally occurring loss- or gain-of-function mutants.
Moreover, loss- or gain-of-function mutants can also be generated
using emerging genome-editing technologies (e.g. CRISPR/Cas9)
(Voytas & Gao, 2014).
4. Data management and analyses
A data management and computational platform for the integra-
tion of large data sets to support predictive biology of complex
systems is necessary to advance CAM research. The US Depart-
ment of Energy Systems Biology Knowledgebase (KBase, http://
kbase.us) provides a computational environment that supports
private and shared research, and is scalable to larger research
communities, such that it could be leveraged by the CAM
community to provide a uniform, narrative-based, computational
platform and reproducible analytical workflows. The centralized
cloud-based platform would be complemented with locally
networked computational resources including a shared LIMS
system, data archiving and retrieval systems, local QA/QC
routines, and analytical workflows. A conceptual design of the
fundamentals needed for development of a CAM computational
platform based on KBase is illustrated in Fig. S2.
IV. Conclusions
The grand challenges caused by ever-increasing human population
and predicted global warming will require scientific innovations to
guarantee a secure and sustainable supply of food, feed, fiber, and
fuel. As a proven mechanism for increasing WUE in plants, CAM
offers great potential for enhancing the sustainable production of
food and biomass on semi-arid, abandoned, or marginal agricul-
tural lands. Thus, CAM research is poised to become a prominent
research area in the plant sciences. The important research

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 499
questions and infrastructure needs discussed earlier could serve as a
reference to prioritize the efforts of the CAM research community.
These critical needs and future opportunities are summarized as a
research roadmap for the CAM community (Fig. 6). The study of
CAM evolution (Fig. 6a) will help elucidate whether the same gene
orthologs have been recruited to a CAM function in different
species, providing guidance for CAM gene discovery using a
systems biology approach (Fig. 6b). For example, if the same gene
orthologs have been recruited to a CAM function, the discovery of
essential CAM genes and cis-regulatory changes required for CAM
can be expedited through comparative analysis of omics data
obtained from multiple diverse CAM species with C3 species as
comparators (Figs 2, 3, 6c). This comparative analysis could
identify four types of CAM genes: (1) CAM genes that have
functionally equivalent C3 gene orthologs without significant
differences in developmental, temporal, or stress-responsive
expression patterns between CAM and C3 species; (2) CAM
genes that have functionally equivalent C3 gene orthologs with
significant differences in developmental, temporal, or stress-
responsive expression patterns between CAM and C3 species; (3)
CAM genes that have orthologs in C3 species but have gained new
function; and (4) CAM-specific genes that have no orthologs in C3
species. Knowledge about these four types of CAM-related genes
could inform the best strategy for CAM-into-C3 engineering
(Figs 4, 6d) to enable enhanced food and bioenergy production on
drylands using existing C3 crops (Fig. 6e). Specifically, CAM
engineering will likely require the transfer of the type 2, 3 and 4
CAM genes described earlier, along with the cis-regulatory changes
required to ensure CAM-like gene expression, into C3 species, and
their integration with the C3 gene orthologs of the type 1 CAM
genes, to form a complete CAM system. However, reiterative
rounds of engineering the introduced genes will likely be necessary
CAM research roadmap
(a) CAM diversity and evolution
C. minor
C. pratensis
Clusia tocuchensis
Portulaca oleracea
Phalaenopsis 
equestris
E. crista-galli
Erycina pusilla
(g)
CAM germplasm (d)
CAM biodesign
(e) C3 or C4 crops with CAM
(f) CAM crop production
CAM systems biology(b)
Sedum
Opuntia
Pineapple
KalanchoeAgave
Ice plant
(c)
Computational
biology
Fig. 6 Crassulacean acidmetabolism (CAM) research roadmap. (a) Genetic diversity and evolution of CAM. (b) CAM systems biologywith a focus on genomics,
proteomics, andmetabolomics, building image©Mayrum,antenna image©2008Angelhell,molecule images courtesyof theEuropeanBioinformatics Instituteat
http://www.ebi.ac.uk/training/online/course/introduction-metabolomics/what-metabolomics, Opuntia image © 2011 Li Jingwang, pineapple image © 2014
Denyshuter. (c) Data management and analysis, building image © 2011 Chuvipro, network image © 2014 aleksandarvelasevic. (d) CAM engineering, building
image © 2014 Bilgic. (e) C3 or C4 crops engineered with CAM for improvement in water-use efficiency and drought tolerance, cotton image © 2010
EIFIacodelNorte, poplar image©2008DNT59. (f) CAMcropproduction,Agave image©Noradoa, pineapple image© 2011 sybil,Opuntia image© 2012Pgiam;
and (g) CAM plant germplasm collection, glasshouse image © 2011 VLADGRIN, people images © Rawpixel, Agave images © 2009 Magnolja.
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist500
to optimize the performance of the CAM system, particularly in
the case of an inducible CAM pathway analogous to those present
in facultative CAM species. Furthermore, the silencing of
endogenous C3 gene orthologs of the type 2 and 3 CAM genes
might be useful to avoid potential conflicts between CAM and C3
in engineered plants; however, the existence of facultative CAM
species suggests that such conflicts are unlikely or can be overcome
readily. In addition, deep understanding of the molecular basis of
adaptive evolution of CAM plants could provide knowledge for
informing genetic improvement in CAM crops for food, feed, and
bioenergy production (Fig. 6f). For example, comparative ge-
nomics analysis of various Agave species in the CAM germplasm
collection (Fig. 6g) could provide molecular information for
increasing cold tolerance in cold-sensitive Agave crop species such
as A. tequilana. Similarly, comparative genomics analysis of
various pineapple and cactus species in the CAM germplasm
collection (Fig. 6g) could provide molecular information for
genetic improvement of food and feed quality and biomass yield in
these two CAM crop species through genomic selection and
breeding.
Acknowledgements
This review is based on work supported by the Department of
Energy (DOE), Office of Science, Genomic Science Program
under award number DE-SC0008834, and the National Science
Foundation under award number DEB-1252901. The contents of
this review are solely the responsibility of the authors and do not
necessarily represent the official views of the DOE. Additional
support is provided by grants from the Nevada Agricultural
Experiment Station (project numbersNAES 000377 and 000380),
and the UK Biotechnology and Biological Sciences Research
Council (grant no. BB/F009313/1). The authors wish to thank
Mary AnnCushman for critical review and clarifying comments on
the manuscript and Lori Kunder (Kunder Design Studio) for
assistance with figure preparation. Oak Ridge National Laboratory
is managed by UT-Battelle, LLC for the US DOE under Contract
Number DE–AC05–00OR22725.
Xiaohan Yang1*, John C. Cushman2, Anne M. Borland1,3,
Erika J. Edwards4, Stan D. Wullschleger5, Gerald A. Tuskan1,
Nick A. Owen6, Howard Griffiths6, J. Andrew C. Smith7,
Henrique C. De Paoli1, David J. Weston1, Robert
Cottingham1, James Hartwell8, Sarah C. Davis9, Katia
Silvera10, Ray Ming11,12, Karen Schlauch13, Paul Abraham14,
J. Ryan Stewart15, Hao-Bo Guo16, Rebecca Albion2,
Jungmin Ha2, Sung Don Lim2, Bernard W. M. Wone2, Won
Cheol Yim2, Travis Garcia2, Jesse A. Mayer2, Juli Petereit13,
Sujithkumar S. Nair5, Erin Casey3, Robert L. Hettich14,
Johan Ceusters17, Priya Ranjan1, Kaitlin J. Palla1,
Hengfu Yin18, Casandra Reyes-Garcı´a19, Jose´ Luis Andrade19,
Luciano Freschi20, Juan D. Beltra´n7, Louisa V. Dever8,
Susanna F. Boxall8, Jade Waller8, Jack Davies8, Phaitun
Bupphada8, Nirja Kadu8, Klaus Winter10, Rowan F. Sage21,
Cristobal N. Aguilar22, Jeremy Schmutz23,24, Jerry Jenkins23
and Joseph A. M. Holtum25
1Biosciences Division, Oak Ridge National Laboratory, Oak
Ridge, TN 37831-6407, USA;
2Department of Biochemistry and Molecular Biology, University
of Nevada, MS330, Reno, NV 89557-0330, USA;
3School of Biology, Newcastle University, Newcastle upon Tyne,
NE1 7RU, UK;
4Department of Ecology and Evolutionary Biology, Brown
University, Box G-W, Providence, RI 02912, USA;
5Environmental Sciences Division, Oak Ridge National
Laboratory, Oak Ridge, TN 37831-6301, USA;
6Department of Plant Sciences, University of Cambridge,
Downing Street, Cambridge, CB2 3EA, UK;
7Department of Plant Sciences, University of Oxford, South Parks
Road, Oxford, OX1 3RB, UK;
8Department of Plant Sciences, Institute of Integrative Biology,
University of Liverpool, Liverpool, L69 7ZB, UK;
9Voinovich School of Leadership and Public Affairs and Depart-
ment of Environmental and Plant Biology, Ohio University,
Athens, OH 45701, USA;
10Smithsonian Tropical Research Institute, PO Box 0843-03092,
Balboa, Ancon, Republic of Panama;
11Department of Plant Biology, University of Illinois at Urbana-
Champaign, Urbana, IL 61801, USA;
12FAFU and UIUC-SIB Joint Center for Genomics and
Biotechnology, Fujian Agriculture and Forestry University,
Fuzhou 350002, China;
13Nevada Center for Bioinformatics, University of Nevada,
MS330, Reno, NV 89557-0330, USA;
14Chemical Sciences Division, Oak Ridge National Laboratory,
Oak Ridge, TN 37831, USA;
15Department of Plant and Wildlife Sciences, Brigham Young
University, 4105 Life Sciences Building, Provo, UT 84602, USA;
16Department of Biochemistry and Cellular and Molecular
Biology, University of Tennessee, Knoxville, TN 37996, USA;
17Department of M²S, Faculty of Engineering Technology, TC
Bioengineering Technology, KU Leuven, Campus Geel,
Kleinhoefstraat 4, B-2440, Geel, Belgium;
18Key Laboratory of Forest Genetics and Breeding, Research
Institute of Subtropical Forestry, Chinese Academy of Forestry,
Fuyang, 311400, China;
19Centro de Investigacio´nCientı´fica de Yucata´n, Calle 43No. 130,
Colonia Chuburna´ de Hidalgo, CP 97200, Me´rida, Me´xico;
20Department of Botany, University of Sa˜o Paulo, Sa˜o Paulo
05508-090, Brazil;
21Department of Ecology and Evolutionary Biology, University of
Toronto, Toronto, ON M5S3B2, Canada;
22Department of Food Research, School of Chemistry,
Universidad Auto´noma de Coahuila, Saltillo, Me´xico;
23HudsonAlpha Institute for Biotechnology, 601 Genome Way,
Huntsville, AL 35801, USA;
24US Department of Energy Joint Genome Institute, 2800
Mitchell Drive, Walnut Creek, CA 94598, USA;
25College of Marine and Environmental Sciences, James Cook
University, Townsville, 4811, QLD Australia
(*Author for correspondence: tel +1 865 241 6895;
email yangx@ornl.gov)

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 501
References
Aida R, Shibata M. 1996. Transformation of Kalanchoe blossfeldiana mediated by
Agrobacterium tumefaciens and transgene silencing. Plant Science 121: 175–185.
Arakaki M, Christin P-A, Nyffeler R, Lendel A, Eggli U, Ogburn RM, Spriggs E,
Moore MJ, Edwards EJ. 2011. Contemporaneous and recent radiations of the
world’s major succulent plant lineages. Proceedings of the National Academy of
Sciences, USA 108: 8379–8384.
Borland AM. 1996.Amodel for the partitioning of photosynthetically fixed carbon
during the C3–CAM transition in Sedum telephium. New Phytologist 134: 433–
444.
BorlandAM,GriffithsH,Hartwell J, Smith JAC. 2009.Exploiting the potential of
plants with crassulacean acid metabolism for bioenergy production on marginal
lands. Journal of Experimental Botany 60: 2879–2896.
BorlandAM,Hartwell J,WestonD, SchlauchK, Tschaplinski T, TuskanG, Yang
X, Cushman JC. 2014. Engineering crassulacean acid metabolism to improve
water-use efficiency. Trends in Plant Science 19: 327–338.
Borland AM, Yang X. 2013. Informing the improvement and biodesign of
crassulacean acid metabolism via system dynamics modelling. New Phytologist
200: 946–949.
Boxall SF, Foster JM, Bohnert HJ, Cushman JC, Nimmo HG, Hartwell J. 2005.
Conservation and divergence of circadian clock operation in a stress-inducible
crassulacean acid metabolism species reveals clock compensation against stress.
Plant Physiology 137: 969–982.
Chao Y-E, Zhang M, Feng Y, Yang X-E, Islam E. 2010. cDNA-AFLP analysis of
inducible gene expression in zinc hyperaccumulator Sedum alfredii Hance under
zinc induction. Environmental and Experimental Botany 68: 107–112.
CheungCM, PoolmanMG, Fell DA, Ratcliffe RG, Sweetlove LJ. 2014.Adiel flux
balance model captures interactions between light and dark metabolism during
day–night cycles in C3 and crassulacean acid metabolism leaves. Plant Physiology
165: 917–929.
Christin P-A, ArakakiM,OsborneCP, Br€autigamA, Sage RF,Hibberd JM, Kelly
S, Covshoff S, Wong GK-S, Hancock L et al. 2014. Shared origins of a key
enzymeduring the evolutionofC4 andCAMmetabolism. Journal of Experimental
Botany 65: 3609–3621.
Christin P-A, Arakaki M, Osborne CP, Edwards EJ. 2015. Genetic enablers
underlying the clustered origins of C4 photosynthesis in angiosperms. Molecular
Biology and Evolution. doi: 10.1093/molbev/msu410.
Colunga-Garc
ıaMar
ın P, Zizumbo-Villarreal D. 2007. Tequila and other Agave
spirits from west-central Mexico: current germplasm diversity, conservation and
origin. Biodiversity and Conservation 16: 1653–1667.
Cook B, Smerdon J, Seager R, Coats S. 2014. Global warming and 21st century
drying. Climate Dynamics 43: 1–21.
Crayn DM, Winter K, Schulte K, Smith JAC. 2015. Photosynthetic pathways in
Bromeliaceae: phylogenetic and ecological significance of CAM and C3 based on
carbon-isotope ratios for 1893 species. Botanical Journal of the Linnean Society.
doi: 10.1111/boj.12275.
Crayn DM, Winter K, Smith JAC. 2004. Multiple origins of crassulacean acid
metabolism and the epiphytic habit in the Neotropical family Bromeliaceae.
Proceedings of the National Academy of Sciences, USA 101: 3703–3708.
Cushman JC, Agarie S, Albion R, Elliot S, Taybi T, Borland A. 2008a. Isolation
and characterization ofmutants of common ice plant deficient in crassulacean acid
metabolism. Plant Physiology 147: 228–238.
Cushman JC,BohnertHJ. 1999.Crassulacean acidmetabolism:molecular genetics.
Annual Review of Plant Physiology and Plant Molecular Biology 50: 305–332.
Cushman JC, Davis SC, Yang X, Borland AM. 2015. Development and use of
bioenergy feedstocks for semi-arid and arid lands. Journal of Experimental Botany.
doi: 10.1093/jxb/erv087.
Cushman JC, Tillett R,Wood J, Branco J, Schlauch K. 2008b. Large-scale mRNA
expression profiling in the common ice plant, Mesembryanthemum crystallinum,
performing C3 photosynthesis and crassulacean acidmetabolism (CAM). Journal
of Experimental Botany 59: 1875–1894.
Dai A. 2013. Increasing drought under global warming in observations andmodels.
Nature Climate Change 3: 52–58.
Davis SC. 2013. Agave: a potential bioenergy feedstock? C4 + CAM Plant Biology
Meeting. 6–9 August 2013, Urbana, IL, USA. [WWW document] URL http://
conferences.igb.illinois.edu/c4cam/ [accessed 23 March 2015].
Davis SC, Dohleman F, Long S. 2011. The global potential for Agave as a biofuel
feedstock. GCB Bioenergy 3: 68–78.
Davis SC, LeBauer DS, Long SP. 2014. Light to liquid fuel: theoretical and
realized energy conversion efficiency of plants using Crassulacean Acid
Metabolism (CAM) in arid conditions. Journal of Experimental Botany 65:
3471–3478.
Davis SC, Ming R, LeBauer DS, Long S. 2015. Toward system-level analysis of
agricultural production from crassulacean acid metabolism (CAM): scaling from
cell to full commercial production. New Phytologist. doi: 10.1111/nph.13522.
DePaoli HC, Borland AM, Tuskan GA, Cushman JC, Yang X. 2014. Synthetic
biology as it relates toCAMphotosynthesis: challenges and opportunities. Journal
of Experimental Botany 65: 3381–3393.
Dever LV, Boxall SF, Knerov
a J, Hartwell J. 2015. Transgenic perturbation of the
decarboxylation phase of Crassulacean acid metabolism alters physiology
and metabolism but has only a small effect on growth. Plant Physiology 167:
44–59.
Eddy SR. 2001. Non-coding RNA genes and the modern RNA world. Nature
Reviews Genetics 2: 919–929.
Edwards EJ, OgburnRM. 2012.Angiosperm responses to a low-CO2 world: CAM
andC4photosynthesis as parallel evolutionary trajectories. International Journal of
Plant Sciences 173: 724–733.
Feugang J, Konarski P, Zou D, Stintzing F, Zou C. 2006. Nutritional and
medicinal use of Cactus pear (Opuntia spp.) cladodes and fruits. Frontiers in
Bioscience 11: 2574–2589.
Freschi L, Mercier H. 2012. Connecting environmental stimuli and crassulacean
acid metabolism expression: phytohormones and other signaling molecules.
Progress in Botany 73: 231–255.
Freschi L, RodriguesMA,DominguesDS, Purgatto E, Van SluysM-A,Magalhaes
JR, KaiserWM,MercierH. 2010.Nitric oxidemediates the hormonal control of
crassulacean acid metabolism expression in young pineapple plants. Plant
Physiology 152: 1971–1985.
Gao J, Sun L, Yang X, Liu J-X. 2013. Transcriptomic analysis of cadmium stress
response in the heavymetal hyperaccumulator Sedum alfrediiHance.PLoSOne 8:
e64643.
Garces HMP, Champagne CEM, Townsley BT, Park S, Malho R, Pedroso MC,
Harada JJ, Sinha NR. 2007. Evolution of asexual reproduction in leaves of the
genus Kalanchoe. Proceedings of the National Academy of Sciences, USA 104:
15578–15583.
Garcia de Cort
azar V, Nobel PS. 1990. Worldwide environmental productivity
indices and yield predictions for a CAM plant, Opuntia ficus-indica, including
effects of doubled CO2 levels. Agricultural and Forest Meteorology 49:
261–279.
Garcia-Sogo B, PinedaB, Castelblanque L, AntonT,MedinaM,Roque E, Torresi
C, Beltran JP,MorenoV,Canas LA. 2010.Efficient transformation ofKalanchoe
blossfeldiana and production of male-sterile plants by engineered anther ablation.
Plant Cell Reports 29: 61–77.
GerlandP,RafteryA, SevcikovaH,LiN,GuD,SpoorenbergT,AlkemaL, Fosdick
B, Chunn J, Lalic N et al. 2014. World population stabilization unlikely this
century. Science 346: 234–237.
Givnish TJ, Barfuss MHJ, Ee BV, Riina R, Schulte K, Horres R, Gonsiska PA,
Jabaily RS, Crayn DM, Smith JAC et al. 2011. Phylogeny, adaptive radiation,
and historical biogeography in Bromeliaceae: insights from an eight-locus plastid
phylogeny. American Journal of Botany 98: 872–895.
Givnish TJ, Barfuss MHJ, Van Ee B, Riina R, Schulte K, Horres R, Gonsiska PA,
JabailyRS,CraynDM,Smith JAC et al.2014.Adaptive radiation, correlated and
contingent evolution, and net species diversification in Bromeliaceae. Molecular
Phylogenetics and Evolution 71: 55–78.
Groenhof AC, Smirnoff N, Bryant JA. 1990. The appearance of a new
molecular species of phosphoenolpyruvate carboxylase (PEPC) and the rapid
induction of CAM in Sedum telephium L. Plant, Cell & Environment 13: 437–
445.
Guye P, Li Y, Wroblewska L, Duportet X, Weiss R. 2013. Rapid, modular and
reliable construction of complex mammalian gene circuits. Nucleic Acid Research
41: e156.
Hartwell J. 2006.The circadian clock in CAMplants. In:Hall AJW,McWattersH,
eds. Annual plant reviews: endogenous plant rhythms. Oxford, UK: Blackwell
Publishing, 211–236.
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist502
Hartwell J, Nimmo G, Wilkins M, Jenkins G, Nimmo H. 1999.
Phosphoenolpyruvate carboxylase kinase is a novel protein kinase regulated at the
level of expression. Plant Journal 20: 333–342.
Hibberd JM, Covshoff S. 2010. The regulation of gene expression required for C4
photosynthesis. Annual Review of Plant Biology 61: 181–207.
Hirsch CN, Foerster JM, Johnson JM, Sekhon RS, Muttoni G, Vaillancourt B,
Pe~nagaricano F, Lindquist E, Pedraza MA, Barry K. 2014. Insights into the
maize pan-genome and pan-transcriptome. Plant Cell 26: 121–135.
Holtum JAM, Chambers D. 2010. Feasibility of Agave as a feedstock for biofuel
production in Australia. Canberra, ACT, Australia: Rural Industry Research and
Development Corporation.
Holtum JAM, Smith JAC, Neuhaus HE. 2005. Intracellular transport and
pathways of carbon flow in plants with crassulacean acid metabolism. Functional
Plant Biology 32: 429–449.
Holtum JAM,Winter K. 1982. Activities of enzymes of carbonmetabolism during
the inductionof crassulacean acidmetabolism inMesembryanthemumcrystallinum
L. Planta 155: 8–16.
Horn JW, Xi Z, Riina R, Peirson JA, Yang Y, Dorsey BL, Berry PE, Davis CC,
WurdackKJ. 2014.Evolutionary bursts in Euphorbia (Euphorbiaceae) are linked
with photosynthetic pathway. Evolution 68: 3485–3504.
KochKE,KennedyRA. 1980.Characteristics of crassulacean acidmetabolism in the
succulent C4 dicot, Portulaca oleracea L. Plant Physiology 65: 193–197.
Koch KE, Kennedy RA. 1982. Crassulacean acid metabolism in the succulent C4
dicot, Portulaca oleracea L under natural environmental conditions. Plant
Physiology 69: 757–761.
Kore-eda S, Nozawa A, Okada Y, Takashi K, Azad M, Ohnishi J, Nishiyama Y,
Tozawa Y. 2013. Characterization of the plastidic phosphate translocators in the
inducible crassulacean acid metabolism plant Mesembryanthemum crystallinum.
Bioscience Biotechnology and Biochemistry 77: 1511–1516.
Krouk G, Lingeman J, Colon AM, Coruzzi G, Shasha D. 2013. Gene regulatory
networks in plants: learning causality from time andperturbation.GenomeBiology
14: 123.
Le Hou
erou H. 1996. The role of cacti (Opuntia spp.) in erosion control, land
reclamation, rehabilitation and agricultural development in the Mediterranean
Basin. Journal of Arid Environments 33: 135–159.
Lewis SM, Gross S, Visel A, Kelly M, Morrow W. 2015. Fuzzy GIS-based multi-
criteria evaluation for US Agave production as a bioenergy feedstock. GCB
Bioenergy 7: 84–99.
Li HJ, Pattathil S, FostonMB, Ding SY, Kumar R, Gao XD,Mittal A, Yarbrough
JM,HimmelME, Ragauskas AJ et al. 2014. Agave proves to be a low recalcitrant
lignocellulosic feedstock for biofuels production on semi-arid lands.Biotechnology
for Biofuels 7: 50.
Li P, Ponnala L, Gandotra N, Wang L, Si Y, Tausta SL, Kebrom TH, Provart N,
Patel R, Myers CR et al. 2010. The developmental dynamics of the maize leaf
transcriptome. Nature Genetics 42: 1060–1067.
L€uttgeU. 2006.Photosynthetic flexibility and ecophysiological plasticity: questions
and lessons from Clusia, the only CAM tree, in the neotropics. New Phytologist
171: 7–25.
L€uttge U. 2008.Clusia: Holy Grail and enigma. Journal of Experimental Botany 59:
1503–1514.
Marbach D, Costello JC, K€uffner R, Vega NM, Prill RJ, Camacho DM, Allison
KR,KellisM,Collins JJ, StolovitzkyG. 2012.Wisdomof crowds for robust gene
network inference. Nature Methods 9: 796–804.
McBride AC, Dale VH, Baskaran LM, DowningME, Eaton LM, Efroymson RA,
Garten CT Jr, Kline KL, Jager HI, Mulholland PJ. 2011. Indicators to support
environmental sustainability of bioenergy systems.Ecological Indicators11: 1277–
1289.
Michelet L, Zaffagnini M, Morisse S, Sparla F, P
erez-P
erez ME, Francia F,
Danon A, Marchand CH, Fermani S, Trost P. 2013. Redox regulation of the
Calvin–Benson cycle: something old, something new. Frontiers in Plant Science 4:
470.
Middleton AM, Farcot E, Owen MR, Vernoux T. 2012. Modeling regulatory
networks to understand plant development: small is beautiful. Plant Cell 24:
3876–3891.
Miguez FE, Maughan M, Bollero GA, Long SP. 2012. Modeling spatial and
dynamic variation in growth, yield, and yield stability of the bioenergy crops
Miscanthus9 giganteus and Panicum virgatum across the conterminous United
States. GCB Bioenergy 4: 509–520.
MoghaddamMRB,Van den EndeW. 2013. Sweet immunity in the plant circadian
regulatory network. Journal of Experimental Botany 64: 1439–1449.
Monson RK. 2003. Gene duplication, neofunctionalization, and the evolution of
C4 photosynthesis. International Journal of Plant Sciences 164: S43–S54.
Moßhammer M, Stintzing F, Carle R. 2006. Cactus pear fruits (Opuntia spp.): a
review of processing technologies and current uses. Journal of the Professional
Association for Cactus Development 8: 1–25.
Nair SS,KangS,ZhangX,MiguezFE, IzaurraldeRC,PostWM,DietzeMC,Lynd
LR, Wullschleger SD. 2012. Bioenergy crop models: descriptions, data
requirements, and future challenges. GCB Bioenergy 4: 620–633.
Nava-Cruz N, Medina-Morales M, Martinez J, Rodriguez R, Aguilar C.
2014. Agave biotechnology: an overview. Critical Reviews in Biotechnology 24:
1–14.
Nobel PS. 1984. Productivity of Agave deserti: measurement by dry weight and
monthly prediction using physiological responses to environmental parameters.
Oecologia 64: 1–7.
Nobel PS. 1994. Remarkable Agaves and cacti. New York, NY, USA: Oxford
University Press.
Nunez HM, Rodriguez LF, Khanna M. 2011. Agave for tequila and biofuels: an
economic assessment and potential opportunities. GCB Bioenergy 3: 43–57.
OwenNA,GriffithsH. 2013.A systemdynamicsmodel integrating physiology and
biochemical regulation predicts extent of crassulacean acid metabolism (CAM)
phases. New Phytologist 200: 1116–1131.
Owen NA, Griffiths H. 2014. Marginal land bioethanol yield potential of four
crassulacean acid metabolism candidates (Agave fourcroydes, Agave salmiana, Agave
tequilana and Opuntia ficus-indica) in Australia. GCB Bioenergy 6: 687–703.
Raftery AE, Alkema L, Gerland P. 2014. Bayesian population projections for the
United Nations. Statistical Science 29: 58–68.
Raftery AE, Chunn JL, Gerland P, Sevc
ıkov
a H. 2013. Bayesian
probabilistic projections of life expectancy for all countries. Demography 50:
777–801.
Raftery AE, LiN, SevcikovaH,GerlandP,HeiligGK. 2012.Bayesian probabilistic
population projections for all countries. Proceedings of the National Academy of
Sciences, USA 109: 13915–13921.
Russell C, Felker P. 1987. The prickly-pears (Opuntia spp., Cactaceae): a
source of human and animal food in semiarid regions. Economic Botany 41:
433–445.
She W, Lin W, Zhu Y, Chen Y, Jin W, Yang Y, Han N, Bian H, Zhu M, Wang J.
2010. The gypsy insulator of Drosophila melanogaster, together with its binding
protein suppressor of hairy-wing, facilitate high and precise expression of
transgenes in Arabidopsis thaliana. Genetics 185: 1141–1150.
Silvera K, Santiago LS, Cushman JC, Winter K. 2009. Crassulacean acid
metabolism and epiphytism linked to adaptive radiations in the Orchidaceae.
Plant Physiology 149: 1838–1847.
Silvera K, Santiago LS, Cushman JC, Winter K. 2010. The incidence of
crassulacean acidmetabolism inOrchidaceae derived fromcarbon isotope ratios: a
checklist of the flora of Panama and Costa Rica. Botanical Journal of the Linnean
Society 163: 194–222.
Silvestro D, Zizka G, Schulte K. 2014. Disentangling the effects of key
innovations on the diversification of Bromelioideae (Bromeliaceae). Evolution
68: 163–175.
Tam GHF, Chang C, Hung YS. 2013. Gene regulatory network discovery using
pairwise Granger causality. IET Systems Biology 7: 195–204.
Taybi T, Patil S, Chollet R, Cushman J. 2000. A minimal Ser/Thr protein
kinase circadianly regulates phosphoenolpyruvate carboxylase activity in
CAM-induced leaves of Mesembryanthemum crystallinum. Plant Physiology
123: 1471–1482.
Torella JP, Boehm CR, Lienert F, Chen J-H, Wau KC, Silver PA. 2014. Rapid
construction of insulated genetic circuits via synthetic sequence-guided
isothermal assembly. Nucleic Acid Research 42: 681–689.
US DOE. 2014. Lignocellulosic biomass for advanced biofuels and bioproducts:
Workshop report.USDepartmentof EnergyOffice of Science. [WWWdocument]
URL genomicscience.energy.gov/biofuels/lignocellulose/ [accessed 23 March
2015].

 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trust
New Phytologist (2015) 207: 491–504
www.newphytologist.com
New
Phytologist Viewpoints Forum 503
Voytas DF, Gao C. 2014. Precision genome engineering and agriculture:
opportunities and regulatory challenges. PLoS Biology 12: e1001877.
Wallach R, Da-Costa N, Raviv M, Moshelion M. 2010. Development of
synchronized, autonomous, and self-regulated oscillations in transpiration rate of
awhole tomato plant underwater stress. Journal of Experimental Botany61: 3439–
3449.
West AG, Gaszner M, Felsenfeld G. 2002. Insulators: many functions, many
mechanisms. Genes & Development 16: 271–288.
WinterK, Foster JG,EdwardsGE,HoltumJAM.1982. Intracellular localization of
enzymes of carbonmetabolism inMesembryanthemum crystallinum exhibiting C3
photosynthetic characteristics or performing crassulacean acid metabolism. Plant
Physiology 69: 300–307.
Winter K,Holtum JAM. 2002.How closely do the d13C values of crassulacean acid
metabolismplants reflect the proportionofCO2fixedduring day andnight?Plant
Physiology 129: 1843–1851.
WinterK,Holtum JAM. 2005.The effects of salinity, crassulacean acidmetabolism
and plant age on the carbon isotope composition of Mesembryanthemum
crystallinum L., a halophytic C3–CAM species. Planta 222: 201–209.
Winter K, Holtum JAM. 2007. Environment or development? Lifetime net CO2
exchange and control of the expression of crassulacean acid metabolism in
Mesembryanthemum crystallinum. Plant Physiology 143: 98–107.
Winter K, Holtum JAM. 2014. Facultative crassulacean acid metabolism (CAM)
plants: powerful tools for unravelling the functional elements of CAM
photosynthesis. Journal of Experimental Botany 65: 3425–3441.
Winter K, Holtum JAM, Smith JAC. 2015.CAMphotosynthesis: a continuous or
discrete trait? New Phytologist. doi: 10.1111/nph.13446.
Yan X, Tan DKY, Inderwildi OR, Smith JAC, King DA. 2011. Life cycle energy
and greenhouse gas analysis for Agave-derived bioethanol. Energy &
Environmental Science 4: 3110–3121.
Supporting Information
Additional supporting information may be found in the online
version of this article.
Fig. S1 Predicted productivity of Agave tequilana and Opuntia
ficus-indica in North America under current and future climate
conditions.
Fig. S2 Data management and analyses for crassulacean acid
metabolism (CAM) research based on the Department of Energy
(DOE) KBase.
Table S1 A list of selected crassulacean acid metabolism (CAM)
crops and information related to their productivity, economic
value, and natural products
Table S2 The hypothesized minimal gene sets for the different
crassulacean acid metabolism (CAM) modules
Table S3 Transcriptomic and genomic information about major
crassulacean acid metabolism (CAM) model species
Please note: Wiley Blackwell are not responsible for the content or
functionality of any supporting information supplied by the
authors. Any queries (other than missing material) should be
directed to the New Phytologist Central Office.
Key words: bioenergy, crassulacean acid metabolism (CAM), drought, genomics,
photosynthesis, roadmap, synthetic biology, water-use efficiency (WUE).
www.newphytologist.com
www.newphytologist.com
np-centraloffice@lancaster.ac.uk
np-usaoffice@lancaster.ac.uk
27
New Phytologist (2015) 207: 491–504 
 2015 ORNL/UT-Battelle
New Phytologist
 2015 New Phytologist Trustwww.newphytologist.com
ViewpointsForum
New
Phytologist504