T cell adaptive immunity proceeds through environment-induced adaptation from the exposure of cryptic genetic variation

8/3/2019 T cell adaptive immunity proceeds through environment-induced adaptation from the exposure of cryptic genetic v

1/27

1

1

T cell adaptive immunity proceeds through environment-2

induced adaptation from the exposure of cryptic genetic3

variation4

5

James M. Whitacre*, Joseph Lin, and Angus Harding6

7

James M. Whitacre8

University of Birmingham9

Department of Computer Science10

Edgbaston, B15 2TT, UK11

[email protected]

13

Joseph Lin14

Sonoma State University15

Department of Biology16

1801 East Cotati Ave.17

Rohnert Park, CA 94928, USA18

19

Angus Harding20

The University of Queensland Diamantina Institute21


2/27

2

Princess Alexandra Hospital22

Ipswich Road23

Woolloongabba, QLD 4102, Australia24

25

*To whom correspondence should be addressed26


3/27

3

Summary27

Evolution is often characterized as a process involving incremental genetic changes that are slowly28

discovered and fixed in a population through genetic drift and selection. However, a growing body29

of evidence is finding that changes in the environment frequently induce adaptations that are much30

too rapid to be accounted for by an incremental genetic search process. Rapid evolution in response31

to environmental change is hypothesized to be driven by mutations present within the population32

that are silent or cryptic within the first environment but are co-opted or exapted to the new33

environment, providing a selective advantage once revealed. Although the hypothesis that cryptic34

mutations facilitate evolution was recently confirmed experimentally in RNA enzymes, the role of35

cryptic mutations in facilitating evolution in complex phenotypes has not been proven.36

In this paper, we describe the unambiguous relationships between cryptic genetic variation and37

complex phenotypic responses within the immune system. By reviewing the biology of the adaptive38

immune system through the lens of evolution, we show that T-cell adaptive immunity constitutes an39

exemplary model system where cryptic alleles drive rapid adaptation of complex traits. In naive T40

cells, normally cryptic differences in TCR reveal diversity in activation responses when the cellular41

population is presented with a novel environment during infection. We summarise how the42

adaptive immune response presents a well studied and appropriate experimental system that can be43

used to confirm and expand upon theoretical evolutionary models describing how seemingly small44

and innocuous mutations can drive rapid cellular evolution.45

46

Key Words47

Cryptic genetic variation (CGV), evolution, adaptive immunity, exaptation48

49


4/27

4

50

1. Introduction: The role of cryptic genetic variation (CGV) in evolution51

The environmental exposure of heritable trait differences can take place in populations through a52

phenomena known as cryptic genetic variation (CGV) [1] (for reviews see [2-4]). CGV describes53

standing genetic variation that is phenotypically hidden in populations within their native habitats54

but that is revealed through changes to the environment [2]. Studies using several model organisms55

have reported CGV in populations subjected to stressful environmental conditions such as changes56

in pH, temperature, nutrient concentrations, moisture, salinity, oxygen, altitude, habitat, and57

hormone levels. For instance, in studies ofDrosophila melanogaster, the presence of cross veins in58

wings, wing shape, scutellar bristle number, and the patterning of photoreceptors in the eye are all59

highly polygenic traits that have shown little variation within native environments but considerable60

variation under modified environmental conditions. Through careful experimentation, many of61

these trait differences have been found to have a heritable basis that originates from previously62

cryptic alleles and not from the discovery of novel alleles, (see [2]).63

One reason that CGV could be of scientific interest is that may help to explain observations of a64

quick tempo in evolution that can keep up with rapid environmental and ecological change. Rapid65

evolution has been reported in many species under conditions where local populations are presented66

with novel environments [5-7]. The pace of this evolutionary change is orders of magnitude faster67

than what is expected based on inferences from the fossil record [7] and is unlikely to be driven by68

de novo mutations due to the relatively slow pace that novel alleles are discovered and fixed in a69

population. On the other hand, standing genetic variation could support faster evolution in novel70

environments because it is available immediately at the time that selective conditions change [8].71

Standing variation can also contain multiple copies of a newly beneficial allele and thus is more72

likely to be fixed in a population and in a shorter amount of time [8]. Because cryptic genetic73

variation is selectively neutral in the original environment, it can accumulate and be maintained,74


5/27

5

while its release of heritable trait differences in an environment dependent manner might provide a75

large resource for rapid evolutionary responses to environmental change.76

In line with this paradigm, a recent study by Wagner and colleagues, using a ribozyme RNA77

enzyme as their experimental model, confirmed that ribozyme populations containing cryptic78variation more rapidly adapted to new substrates compared to ribozyme populations that did not79

contain cryptic mutations [9]. This study provides the first direct experimental confirmation that80

cryptic genetic changes can promote rapid evolution. While there are studies that strongly implicate81

CGV in rapid evolution of complex phenotypes (for example see [10, 11]), the direct connection82

between cryptic alleles and novel traits has not been confirmed in any of these studies. This is83

because directly demonstrating the relationship between CGV and rapid evolution in a complex84

organism remains a challenging milestone [4], in part because genetic diversity can be distributed85

over a large genome, traits can have a large but unknown genetic footprint, and several factors such86

as mutations, stochastic effects, and environmental heterogeneity can influence allele frequencies87

and thus confound links between CGV and adaptation.88

To formally confirm the role of cryptic variation in the evolution of complex phenotypes it is89

essential to identify model systems where specific and well-characterized cryptic alleles can be90

shown to drive trait differences with adaptive significance in new environments. In this review we91

explain how the adaptive immune response of T cells provides an ideal model system that clearly92

demonstrates the potentially important relationship between CGV and rapid evolution.93

2. The Adaptive Immune system: a Primer94

The adaptive immune system provides vertebrates with the ability to recognize and respond to a95

variety of pathogens including parasites, bacteria, fungi and viruses. The primary effector cells of96

the adaptive immune system are B cells and T cells, which are derived from the same multi-potent97

hematopoietic stem cells. B cells mature in the bone marrow and are involved primarily in the98


6/27

6

creation of antibodies. B cells express the B cell receptor (BCR), which is a membrane-bound99

antibody molecule. Different BCRs recognise and bind to their respective foreign molecules100

(antigens). Once a B cell binds its cognate antigen and receives an additional signal from a helper T101

cell, it activates proximal signalling events triggering its differentiation into a short lived effector102

plasma cell that functions to secrete antibodies (soluble forms of the BCR) that bind to its cognate103

antigen. Antibodies bound to foreign molecules make them easy targets for phagocytosis and other104

clearance mechanisms. Subpopulations of activated B cells differentiate into long-lived antigen105

specific memory B cells, which remain in circulation and can respond quickly if the host is re-106

infected by the same pathogen.107

T cells mature in the thymus and are involved primarily in cell-mediated immune responses. They108

differ from B cells in that their antigen binding receptor, known as the T cell receptor (TCR) can109

only recognise peptide antigens that have been proteolytically cleaved and presented in the context110

of a host-cell Major Histocompatibility Complex (MHC) protein (Supplemetary Figure S2). All111

nucleated cells are capable of expressing antigen in an MHC class I complex and activating a T cell112

response. However B-cells, dendritic cells and macrophages additionally express MHC class II113

molecules, as well as other immune-stimulatory molecules, that allow enhanced activation of T114

cells, and are classified as professional antigen presenting cells (APCs).115

There are two primary types of T cells, cytotoxic T cells (CTLs) and helper T cells. Naive cytotoxic116

T cells (CTLs) are activated when they recognise their cognate antigenic peptide presented in the117

context of an infected host cells MHC class I molecule. Once activated, the CTL undergoes clonal118

expansion where it simultaneously differentiates into a mature CTL and divides rapidly, generating119

a large number of activated effector cells that travel through the body and scan for infected cells120

expressing the same antigenic peptide. Active CTLs induce lysis and apoptosis in infected cells.121

Most effector CTLs undergo apoptosis and are cleared away by phagocytic cells once the infection122

is resolved, however a subset of CTLs differentiate into memory cells that can rapidly respond to123


7/27

7

subsequent infections by the same pathogen. Helper T cells have no cytotoxic activity, and instead124

function to regulate the adaptive immune response. The TCR of T helper cells recognise antigen125

peptide bound to class II MHC molecules on the host cell surface. The activation of naive T helper126

cells results in differentiation and clonal expansion similar to CTLs. Unlike CTLs, activation of127

helper T-cells triggers them to release cytokines that regulate the activity of many types of cells128

including APCs, B cells, and cytotoxic T-cells.129

Together, the components of the adaptive immune system respond rapidly and specifically to130

infection from previously un-exposed pathogens [12]. Importantly, in most cases, during subsequent131

exposure to the same pathogen, the memory response is so protective that no outward signs of re-132

exposure may be seen [13]. Although many of the characteristics of the adaptive immune response133

were demonstrated decades ago [14, 15], the elegance and efficiency of antigen recognition was not134

fully appreciated until the genetic and molecular basis of this response was determined [16].135

3. The immunological role of the T cell Receptor (TCR)136

Under conditions where foreign proteins are present, such as during a viral infection or137

phagocytosis of a pathogen, foreign peptides become available to be processed and loaded onto138

MHC molecules. It is the replacement of a proportion of endogenous self-peptides by foreign139

peptides that activates the T cells in the periphery. T cell activation involves an elaborate series of140

events between the T cell and the MHC-expressing antigen-presenting cell (APC) (Supplemtary141

Box 2 and Figure S2). Several receptor-ligand interactions between T cells and APCs are important142

for activation, however activation differences between T cells are determined by whether TCR143

recognizes the foreign peptide loaded onto the MHC molecule. Interestingly, T cell activation has144

been shown to be digital in nature: a T cell can only exist in an Off (resting) state or an On145

(activated) state [17-22]. This means that as long as the threshold of activation is achieved,146

relatively weak stimuli (low affinity) will induce the same response as strong stimuli (high affinity)147

at the single cell level. In other words, selectively relevant trait differences in T cells can be148


8/27

8

captured by a single binary variable [18, 22]. Although there are some differences in how activation149

takes place for different classes of T cells and the different types of MHC molecules, TCR150

recognition of the peptide loaded MHC remains the defining step in all cases of T cell activation.151

The genetic basis of the TCR repertoire: The development of a diverse T cell repertoire is a152critical step for maintaining the overall health of an organism as it allows the adaptive immune153

system to respond to many different pathogenic antigens. Genetic variation of the TCR across the T154

cell population provides the requisite variety of cell response options that is necessary for the155

adaptive immune response. These genetic differences among T cells arise through the stochastic156

recombination of a diverse set of gene segments that make up the TCR. The TCR itself is157

composed of several subunits ultimately expressed on the surface of the cell. The and chains of158

the TCR are directly responsible for ligand recognition and therefore represent the variable and159

unique portion of each TCR complex that interacts with peptide-MHC complexes. The other160

subunits that comprise the TCR (collectively called CD3 chains) are non-variable and are161

responsible for propagating signals in the cytoplasm of the cell. These signalling subunits contain162

highly conserved motifs that are critical for initiating the intracellular events leading to T cell163

activation.164

T cell development begins with the entry of T cell progenitors into the thymus, which initiates a165

program of differentiation that includes T cell receptor gene rearrangement and expression (Figure166

1 and Supplementary Box 2). The variability of each and chain is generated by the random167

selection and imprecise joining of different germline encoded gene segments [23]. The168

recombination of these gene segments also involves the addition of random nucleotides at each gene169

segment junction further increasing the variability seen in TCR repertoires. The and chains, if170

recombined successfully, are then co-expressed and assembled with the CD3 chains to form a TCR171

with specificity that is unique to that cell. Studies have suggested that this mechanism can172

potentially generate 1012 to 1015 different possible TCRs [24]. The stochastic generation of the TCR173


9/27

9

by recombination thus constitutes a large, pre-established genotype space for TCR and along with174

its associated antigenic specificity provides a relatively simple and well-defined fitness landscape of175

T cell activation.176

TCR/MHC interactions within the thymus dictate selection of the nave T cell population:177Mature T cells present in the periphery must be able to carefully distinguish between cells178

presenting harmless endogenous self-peptides and cells presenting foreign peptides derived from179

potentially dangerous pathogens. Once the TCR has been assembled through stochastic DNA180

recombination, it is rigorously tested within the thymus to ensure that it is functional but will not181

respond to self-peptides (Figure 1 and Supplementary Box 2). Defects in this selection process can182

lead to either profound immunodeficiencies or debilitating autoimmune diseases. At the heart of183

this selection process is the determination of how the newly formed TCR interacts with a peptide-184

loaded MHC molecule. First, TCRs are screened for their ability to weakly recognize endogenous185

self peptides displayed on an MHC molecule within the cortex of the thymus (Figure 1 and186

reviewed in [25]).187

Those mature T-cells that survive thymic selection represent fewer than 5% of the starting pre-188

selected thymocyte population and are exported from the thymus, entering the circulatory system to189

ultimately reside in the peripheral lymphoid organs. It is believed that following selection in the190

thymus, the range of TCR variation is between 106 to 107 [26].191

4. T cells provide an unambiguous example of evolution driven by environment-192

exposed CGV193

With evidence of cryptic genetic variants being co-opted into beneficial functions [27, 28] (for194example during exaptation [29]), it is believed that there could be many circumstances where novel195

environments reveal CGV and may drive rapid evolution in natural populations [9-11]. In order to196

clearly demonstrate this evolutionary pathway, several conditions must be met. Populations197

typically must occupy habitats that are stable over many generations (CGV condition 1) and198


10/27

10

accumulate considerable genetic diversity while exhibiting only small trait differences (CGV199

condition 2). When presented with novel environments, the same population needs to display new200

quantitative traits (CGV condition 3) that have a heritable basis (CGV condition 4) and thus may201

contribute to evolution within the novel environment [2, 30].202

Each of these conditions corresponds with well characterized features in the adaptive immune203

system. T cells develop in a stable thymic environment where a sub-population is selected to mature204

and leave the thymus based on positive/negative selection. To satisfy the selection conditions, T205

cell receptors must have an affinity to host cell protein fragments that is strong enough to result in206

non-trivial adhesion between the T cell and host cell but is not so strong as to elicit T cell activation.207

The result is a thymic environment that efficiently selects for mature T cells with similar behaviour208

under normal conditions that are devoid of foreign antigens. These selection conditions within the209

thymus are stable and remain the same throughout the lifetime of the organism (CGV condition 1).210

The genome:proteome landscape of TCR contains many distinct proteins that are functionally211

equivalent for the selection conditions of the thymus. Because thymic selection cannot distinguish212

between these neutral variants, this allows for the accumulation of a genetically diverse repertoire213

of receptors across the nave T cell population (CGV condition 2). However, these T cell receptors214

retain functional differences that, while hidden in the initial selection context of the thymus, can215

lead to the expression of highly distinct cell responses (activation) when T-cells are removed from216

their initial developmental environment and presented with antigens. Each novel (antigen)217

environment leads to the activation of a small unique subset of nave T-cells, which at the218

population level appears as a widening of the T-cell populations affinity distribution (Figure 1).219

The antigen-specific exposure of T-cell affinity differences is an example of trait differences220

released under novel environmental conditions and is a hallmark of CGV (CGV condition 3).221

Antigen-specific T-cell response differences have a clear genetic basis that is found in the variation222

of alpha and beta chains (CGV condition 4). Moreover, in a properly-functioning immune system,223


11/27

11

the exposure of response differences leads to activation and subsequent clonal expansion for a small224

selected set of T cells, demonstrating the evolutionary relevance of this CGV. Defining225

characteristics of the innate and adaptive immune response in T cells thus correspond with defining226

characteristics of evolutionarily relevant CGV phenomena (summarised in Figure 2). Together,227

these features formally establish a role for cryptic genetic variation in a well-defined evolving cell228

population.229

An ideal experimental system. T cell adaptive immunity provides ideal genotypic, phenotypic,230

and environmental conditions that allow links between CGV and environment-induced adaptation231

to be clearly established. For one, the adaptive immune response preserves selected phenotypes232

(memory T cells) without continued stimulation from the environment and without the need for233

additional genetic accommodation. CGV can thus be isolated from additional environmental234

changes and genetic changes (such as genetic accommodation and drift) that would continue to take235

place after CGV is revealed in natural populations.236

Adaptive immunity also has a well-characterized genetic basis with a small variable genetic237

footprint, despite T cell activation being a highly polygenic trait. Relevant genetic differences238

between T-cells are constrained to a single TCR receptor with a genetic sequence that is highly239

constrained to a predefined space of recombined alpha/beta chains. TCR is generated through a240

stochastic recombination of gene segments. As a result of this recombination, each gene segment241

can contribute to activation differences in the T cell repertoire yet remains cryptic within a number242

of genetic backgrounds. This provides an important similarity to CGV in sexually reproducing243

populations without the presence of other features of sexual reproduction that would otherwise244

arise, such as linkage, disequilibrium and assortative mating. Furthermore, with the immune system245

there is no need to isolate CGV from other causes of genetic diversity such as environmental246

heterogeneity or density dependent selection.247


12/27

12

The environmental changes that induce T cell responses are also well characterized by changes in248

antigens attached to the surfaces of APCs. Because the time between infections is typically much249

longer than the immune response time (T cell activation and transition to long-lived memory cells),250

environmental volatility is not a confounding factor. In addition, environment-induced trait251

differences are nicely characterized by a digital on/off activation response.252

In contrast, evolution in natural populations proceeds through a concurrent stochastic population253

process where both standing genetic variation and de novo mutations contribute to adaptation, and254

where non-selective factors such as genetic drift and genetic hitchhiking will also influence allele255

frequencies. Studying the contribution of CGV in evolution is also difficult because its role is256

typically revealed in a heterogeneous and dynamic environment where genetic assimilation [31-33]257

and genetic accommodation [34, 35] are sometimes needed to preserve the inheritance of258

phenotypic adaptations in the face of ongoing environmental change [35, 36]. By isolating CGV259

from these factors, T cell adaptive immunity provides an experimental system where the260

contribution of CGV towards rapid evolution of cellular phenotypes can be clearly established and261

confirmed.262

It is worth clarifying why we focus on adaptation in T cells and not B cells. The adaptive response263

in naive T cells arises from CGV exposure within a novel environment and does not involve a264

classic genetic search process where the onset of adaptation would take place by the discovery of a265

novel allele. The story is more complicated in B cells because adaptations can arise from both266

CGV and a genetic search process. In particular, after the environment-dependent activation of B267

cells, a genetic search occurs that involves clonal expansion, hyper-mutation, and selection, with the268

result being a more finely tuned memory B cell. Clearly both gene-induced and environment-269

induced adaptations are therefore relevant to the immune system. By focusing on T cells we remove270

any ambiguity as to the origins of adaptation and also in the subsequent proof of principles that271

were discussed.272


13/27

13

5. Concluding Remarks273

Evolution is traditionally characterized as a process involving incremental genetic changes that are274

discovered and fixed in a population through genetic drift and selection. While the genetic basis of275

heredity is not disputed, it is increasingly recognized that the introduction of novel alleles is not the276

only way that heritable phenotypes originate in species. A growing body of evidence suggests that277

changes in the environment can induce heritable and evolutionarily relevant trait differences that278

drive rapid evolution in populations [2, 3, 8, 11, 31-35].279

However, the typical presence of multiple confounding factors makes it difficult to determine the280

evolutionary significance of cryptic genetic variation [4, 8]. Evidence of environment-exposed281

CGV driving heritable changes in a deme is sporadic at best [4]; it remains difficult to282

systematically uncover, and its plausibility is thus far only indirectly supported by the controlled283

manipulation of artificial environments [8-10] or phylogenetic analysis [8, 11].284

In this article we have recounted how T cell maturation and antigen-specific activation provide a285

clear and unambiguous example of environmental exposure of cryptic genetic variation that drives286

complex phenotypic changes in a cell population. CGV within the TCR repertoire facilitates rapid287

evolutionary responses in the adaptive immune system and is fundamental to the survival of288

vertebrates in the face of infections from diverse pathogenic sources.289

The confirmation and gradual acceptance of CGV-driven evolution amongst the scientific290

community may have broad repercussions to scientific research investigating the preservation of291

species and the eradication of evolvable pathogens. For example, in conservation biology292

intraspecific genetic diversity is a well known factor in extinction risk. However, only genetic293

variants that reveal trait differences under the environmental stresses encountered can mitigate294

extinction risks. It should be possible to quantify GxE interactions for environmental stresses that295

endangered species are expected to encounter with greater frequency and magnitude in the future,296

for instance by selecting stresses based on predictive models of regional climate change. CGV that297


14/27

14

is revealed under such conditions is predicted to contain high evolutionary potential towards future298

adaptive requirements. Compared to overall intraspecific genetic variation, this would provide a299

more adaptively significant biodiversity metric for determining what individuals are most important300

to species conservation efforts. Moreover, with limited conservation budgets, strategies for301

preserving this much smaller CGV footprint may provide a more realistic and realizable302

conservation goal.303

On the other hand, in highly evolvable infections and diseases such as aggressive cancers, the304

accumulation of CGV is likely to play an important role in the rapid evolution of therapy resistance305

that is seen in the latest generation of targeted therapies [37]. Genetic variation is maintained at high306

levels in many tumours and tumours with high genetic diversity have been shown to more rapidly307

evolve resistance to therapies [37, 38]. Interestingly, the therapeutic strategies used today may308

inadvertently select for the evolution of resistance [39]. For instance, therapies are often applied309

long after therapy resistance is first established in the cancer cell population, thereby allowing a310

founder population to grow and acquire the genetic diversity needed for adaptation to the next311

therapy that is eventually applied. Oncologists have yet to consider the value of a dynamic312

therapeutic environment. If the therapy were instead changed at the point of lowest CGV, the cell313

population would be highly vulnerable to new environmental stresses in a manner analogous to the314

extinction risks posed by low genetic diversity in natural populations. CGV will not readily315

accumulate in a rapidly changing environment with strong directional selection and we propose316

therefore that dynamic therapeutic strategies could provide a generic approach to reducing the317

evolutionary potential of cancers containing high levels of genetic diversity.318

Using an ideal model system, this paper formally establishes links between cryptic alleles and319

environment-induced adaptation in a complex cellular phenotype. As confirmation of these320

principles is further documented in other biological contexts and becomes more widely accepted,321


15/27

15

we anticipate that CGV will act as a focal point in efforts to preserve or eradicate evolving322

populations.323


16/27

16

Figure 1Thymic selection of T-cells. A. Immature T-cells (called thymocytes) enter the thymic cortex and, supported by324factors secreted by cortical thymic epithelial cells (cTECs, green), undergo differentiation. During this time TCR re-arrangement325occurs until an active TCR is expressed along with both the CD4 and CD8 co-receptors (Double Positive/DP stage). Those DP326thymocytes that have a functional TCR and survive commit to a single positive (SP) state, (SP thymocytes express either CD4 or327CD8, but not both) and migrate to the thymic medulla. Here, SP thymocytes systematically scan medullary thymic epithelial cells328(mTECs, shown in brown) and dendritic cells presenting self-peptides in their MHC molecules. During the SP thymocyte stage, any329thymocyte that expresses a TCR that has an above threshold affinity to self-peptides undergoes apoptosis and dies. Surviving330thymocytes eventually exit the thymus and enter the periphery as mature, nave T-cells. B. Positive and negative selection is331determined by TCR affinity to their peptide/MHC ligand. Once a TCR is expressed on the cell surface it immediately begins332

interacting with MHC molecules presenting self-peptides. Functional TCRs generate a basal signal output that, although insufficient333 to cross the activation threshold, is still necessary for thymocyte survival. TCRs that cannot recognize peptide/MHC complexes334ultimately die by neglect. Surviving thymocytes then migrate to the thymic medulla where they are exposed to self antigen through335promiscuous gene expression in mTECs and DCs. Any TCR that reacts with a self-peptide with too high of an affinity will induce T336cell activation and apoptosis. Only thymocytes with a functional TCR, which is competent to trigger low-level signalling but does not337cross the activation threshold to self-peptides, is positively selected and permitted to exit the thymus.338

339

Figure 2. Comparing attributes of T cells (cell development and activation) with CGV in natural populations.340In both cases, genetic diversity arises in a stable environment that selects cells/organisms that fall within a narrow band of trait341values. When exposed to a novel environment, trait differences are revealed in the population. In adaptive immunity, we consistently342observe a subset of cells that are selected to undergo clonal expansion. In natural populations, a change in the environment does not343always immediately reveal trait differences that are selectively relevant; however selection of sub-populations with selective344advantage resulting from CGV is expected to take place over sufficiently long timescales.345

346


17/27

17

Acknowledgements347

AH is supported by an NHMRC project grant, an ARC project grant, and The Brain Foundation348

Australia. JW is supported by a DSTO grant and an EPSRC grant (No. EP/E058884/1).349


18/27

18

References350

1. Le Rouzic A., Carlborg . 2008 Evolutionary potential of hidden genetic variation. Trends in Ecology &351Evolution23(1), 33-37.3522. Gibson G., Dworkin I. 2004 Uncovering cryptic genetic variation.Nature Reviews Genetics5(9), 681-690.3533. Schlichting C.D. 2008 Hidden Reaction Norms, Cryptic Genetic Variation, and Evolvability.Annals of the354New York Academy of Sciences1133(1 The Year in Evolutionary Biology 2008), 187-203.3554. McGuigan K., Sgr C.M. 2009 Evolutionary consequences of cryptic genetic variation. Trends in Ecology &356

Evolution24(6), 305-311.357 5. Kinnison M.T., Hendry A.P. 2001 The pace of modern life II: from rates of contemporary microevolution to358pattern and process. Genetica112(1), 145-164.3596. Hendry A.P., Kinnison M.T. 1999 Perspective: the pace of modern life: measuring rates of contemporary360microevolution.Evolution53(6), 1637-1653.3617. Reznick D.N., Ghalambor C.K. 2001 The population ecology of contemporary adaptations: what empirical362studies reveal about the conditions that promote adaptive evolution. Genetica112(1), 183-198.3638. Barrett R.D.H., Schluter D. 2008 Adaptation from standing genetic variation. Trends in Ecology & Evolution36423(1), 38-44.3659. Hayden E.J., Ferrada E., Wagner A. 2011 Cryptic genetic variation promotes rapid evolutionary adaptation in366an RNA enzyme.Nature474, 92-97.36710. McGuigan K., Nishimura N., Currey M., Hurwit D., Cresko W.A. 2011 Cryptic Genetic Variation and Body368Size Evolution in Threespine Stickleback.Evolution65(4), 1203-1211.36911. Palmer A. 2004 Symmetry breaking and the evolution of development. Science306(5697), 828.370

12. Raff M.C. 1973 T and B lymphocytes and immune responses.Nature242(5392), 19.371 13. Sallusto F., Geginat J., Lanzavecchia A. 2004 Central memory and effector memory T cell subsets: function,372generation, and maintenance.Annu Rev Immunol22, 745-763.37314. Paul W.E., Benacerraf B. 1977 Functional specificity of thymus-dependent lymphocytes. Science195(4284),3741293.37515. North R.J. 1969 Cellular kinetics associated with the development of acquired cellular resistance. The Journal376of experimental medicine130(2), 299.37716. Rajewsky K. 1996 Clonal selection and learning in the antibody system.37817. Lin J., Harding A., Giurisato E., Shaw A.S. 2009 KSR1 modulates the sensitivity of mitogen-activated protein379kinase pathway activation in T cells without altering fundamental system outputs.Mol Cell Biol29(8), 2082-2091.380(doi:MCB.01634-08 [pii]38110.1128/MCB.01634-08).38218. Altan-Bonnet G., Germain R.N. 2005 Modeling T cell antigen discrimination based on feedback control of383digital ERK responses. PLoS Biol3(11), e356. (doi:05-PLBI-RA-0208R2 [pii]38410.1371/journal.pbio.0030356).

385 19. Daniels M.A., Teixeiro E., Gill J., Hausmann B., Roubaty D., Holmberg K., Werlen G., Hollander G.A.,386Gascoigne N.R., Palmer E. 2006 Thymic selection threshold defined by compartmentalization of Ras/MAPK signalling.387Nature444(7120), 724-729. (doi:nature05269 [pii]38810.1038/nature05269).38920. Prasad A., Zikherman J., Das J., Roose J.P., Weiss A., Chakraborty A.K. 2009 Origin of the sharp boundary390that discriminates positive and negative selection of thymocytes. Proc Natl Acad Sci U S A106(2), 528-533.391(doi:0805981105 [pii]39210.1073/pnas.0805981105).39321. Das J., Ho M., Zikherman J., Govern C., Yang M., Weiss A., Chakraborty A.K., Roose J.P. 2009 Digital394signaling and hysteresis characterize ras activation in lymphoid cells. Cell136(2), 337-351. (doi:S0092-8674(08)01631-3950 [pii]39610.1016/j.cell.2008.11.051).39722. Chakraborty A.K., Das J., Zikherman J., Yang M., Govern C.C., Ho M., Weiss A., Roose J. 2009 Molecular398origin and functional consequences of digital signaling and hysteresis during Ras activation in lymphocytes. Sci Signal3992(66), pt2. (doi:scisignal.266pt2 [pii]40010.1126/scisignal.266pt2).40123. Schatz D. 2004 V (d) j recombination.Immunological Reviews200(1), 5-11.40224. Davis M.M., Bjorkman P.J. 1988 T-cell antigen receptor genes and T-cell recognition.40325. Starr T., Jameson S., Hogquist K. 2003 Positive and negative selection of T cells.Annual Review of404Immunology21, 139.40526. Arstila T.P., Casrouge A., Baron V., Even J., Kanellopoulos J., Kourilsky P. 1999 A direct estimate of the406human alphabeta T cell receptor diversity. Science286(5441), 958.40727. Waddington C.H. 1942 Canalization of Development and the Inheritance of Acquired Characters.Nature408150(3811), 563.40928. Waddington C.H. 1957 The strategy of the genes, Allen & Unwin London.410


19/27

19

29. Badyaev A.V. 2005 Stress-induced variation in evolution: from behavioural plasticity to genetic assimilation.411Proceedings of the Royal Society B: Biological Sciences272(1566), 877.41230. Whitacre J.M. (In press) Genetic and environment-induced pathways to innovation: on the possibility of a413universal relationship between robustness and adaptation in complex biological systems414(http://www.springerlink.com/content/0577136586542281/)Evol Ecol. (doi:10.1007/s10682-011-9464-z).41531. Waddington C.H. 1953 Genetic Assimilation of an Acquired Character.Evolution7(2), 118-126.41632. Waddington C.H. 1957 The Strategy of the Genes: A Discussion of Some Aspects of Theoretical Biology, Allen417& Unwin.41833. Schmalhausen I., Dobzhansky T. 1949 Factors of evolution: the theory of stabilizing selection, Blakiston Co.419Philadelphia.42034. West-Eberhard M. 2005 Developmental plasticity and the origin of species differences. Proceedings of the421National Academy of Sciences102(Suppl 1), 6543.42235. West-Eberhard M. 2003Developmental plasticity and evolution, Oxford University Press, USA.42336. Pigliucci M., Murren C., Schlichting C. 2006 Phenotypic plasticity and evolution by genetic assimilation.J424Exp Biol209(12), 2362.42537. Tian T., Olson S., Whitacre J.M., Harding A. 2011 The origins of cancer robustness and evolvability.426Integrative Biology3, 17-30. (doi:10.1039/c0ib00046a).42738. Gerlinger M., Swanton C. 2010 How Darwinian models inform therapeutic failure initiated by clonal428heterogeneity in cancer medicine.Br J Cancer103(8), 1139-1143.42939. Read A.F., Day T., Huijben S. 2011 The evolution of drug resistance and the curious orthodoxy of aggressive430chemotherapy. Proceedings of the National Academy of Sciences108(Supplement 2), 10871.431

432433

434


20/27

20

Supplementary Box 1: Cryptic Genetic Variation and Evolution435

Trait robustness towards mutations and environmental variations is conditional upon both the436

genetic and environmental background. Within a particular environment, trait robustness towards437

mutations enables considerable genetic diversity to accumulate under mutation-selection balance.438

While many genetic variants may appear silent or cryptic with this environment, the conditionality439

of trait robustness on the environmental background means that such populations will reveal440heritable trait differences within novel environmental conditions. This phenomena is referred to as441

cryptic genetic variation (CGV) and has been hypothesized to be a source of rapid evolutionary442

adaptation in new environments.443

Confusion about the role of CGV in evolution can be attributed to long-standing difficulties in444

understanding relationships between robustness and adaptation and also due to historical445

descriptions of evolution as a genetic search process. For example, Mutational Robustness the446

stability of the phenotype to mutations was originally predicted to impede evolvability [40, 41]447because it lowers the mutational accessibility of distinct heritable phenotypes from a single448

genotype and it reduces selective differences within a genetically diverse population [41]. Only in449

the last decade has a resolution to this paradox been established for mutational [41-43] and450environmental [44] forms of robustness. For instance, mutational robustness is now believed to451

facilitate evolvability in two ways: first by establishing genotypic neutral networks from which452

large numbers of distinct genotypes can be sampled [41] and second by allowing genetic diversity453

to arise in a population with subsequent trait differences revealed in an environment-dependent454

manner [44]. Both of these pathways to adaptation occur due to the presence of CGV in populations455

(see Figure S3). However, what is not yet well appreciated is the important role that CGV can play456

when evolution occurs in a dynamic environment. In particular, CGV may provide the necessary457

fuel for adaptation under new stressful environments while bypassing the negative selection that458limits phenotypic variability under stable conditions [45].459

460|NN-G|

# of genotypes

connected to NN-G

that are associated

with unique

phenotypes

Genetic Neutral Network

Static Environment

Dynamic Environment


21/27

21

Figure S3 The relationship between cryptic genetic variation and a populations propensity to evolve. Left461Panel: When many genotypes with similar fitness are present in a fitness landscape, these can connect to form a neutral network462(NN, see black nodes). Distinct heritable phenotypes are still accessible through mutations that lead to genotypes that are not463members of NN (see coloured nodes). Larger NN can extend through larger regions of the fitness landscape and have access to a464greater range of distinct heritable phenotypes [41-44]. Right Panel, Genetically-induced adaptation: If the environment is465stable, a population will diffuse and drift over the NN allowing it to sample many heritable phenotypes over time and thereby466supporting evolution. Right Panel, Environmentally-induced adaptation: Both the properties of individual phenotypes and467entire fitness landscapes depend on the environment. In a stable environment, a population is able spread over a region of the neutral468network because these genetic differences are selective cryptic. However a change in the environment can influence each genotype469differently so that genetic distinctions in the population are revealed as phenotypic distinctions as well. At the fitness landscape470level, this appears as a partial collapse of the neutral network.471

Supplementary Box 2: Summary of T cell selection in the thymus472

Thymic development of T cells begins with the entry of T cell progenitors into the thymus through473

blood vessels near the cortico-medullary junction [46]. Thymic entry into the cortex initiates a474

program of differentiation that includes TCR gene rearrangement and expression, in combination475with changes in expression of cell surface proteins [47]. The and chains of the TCR are directly476

responsible for ligand recognition and therefore represent the variable and unique portion of each477

TCR complex that interacts with peptide-MHC complexes. The variability of each and chain is478

generated by the random selection and imprecise joining of different germline encoded gene479segments. During development, thymocytes first begin forming the TCR chain by the assembling480

one variable (V), one diversity (D), and one joining (J) gene segment to an invariant constant region481

(C) from a library of V (n=52), D (n=2), and J (n=13) segments that are all initially encoded482

within each gene locus [23]. The recombination of these gene segments also involves the addition483of random nucleotides at each gene segment junction further increasing the variability seen in TCR484

repertoires. The newly formed chain pairs up with a surrogate, pre-formed chain called the pre-485

T cell (pT), as well as with the CD3 chains, forming a pre-TCR. The pre-TCR has all the486

signalling components of a mature TCR, leading to cellular proliferation, the arrest of chain gene487

re-arrangements, and the expression of CD4 and CD8. These CD4/CD8 double positive (DP)488

thymocytes comprise the majority of thymocytes within the thymus.489


22/27

Once the pre-TCR DP thymocytes c490

proliferating they become smaller a491

the genetic re-arrangement of the T492

chain. The chain rearrangement is493

to that of the chain except that the494

composed of only one V (n= approx495

one J (n=61) segment fused again to496invariable constant region. The a497

chains, if recombined successfully,498

co-expressed and assembled with th499

chains to form a TCR with specifici500

unique to that cell. Studies have su501

that this mechanism can potentially502

1012

to 1015

different possible TCRs503

stochastic generation of the TCR by504

recombination thus constitutes a lar505

established genotype space for TCR506along with its associated antigenic s507

provides a relatively simple and wel508

fitness landscape of T cell activatio509

Expression of the successfully rearr510

TCR together with the TCR co-rece511

leads to the commitment to the TCR512

lineage and the expression of CD4513

co-receptors to create double positi514

thymocytes [48, 49]. DP thymocyte515

expressing TCR

move actively w516 cortical microenvironment, scannin517

cortical epithelial cells (cTECs) MH518

molecules [50]. Functional TCR519

that respond to self peptide-MHC c520

on cTECs are induced to survive an521

proliferate [51], whereas TCRs defe522

MHC binding die from neglect [52]523

selection for functional TCRs withi524

cortex is known as positive selectio525

positive selection, differential kineti526

TCR-MHC interaction drives DP th527

positive (SP) thymocytes [52].528

Positively selected SP thymocytes t529

before being licensed to exit the thy530

medulla is a specialised environmen531

tolerance to self-antigens, including532

epithelial cells (mTECs) ectopically533

22

Figure S2 T cell activation. The TCchains and CD3) and the co-receptors CD4 o

tyrosine kinases of the Src family. When th

presenting its cognate antigen, the co-recethe MHC molecule on the APC surface. Th

kinases by the phosphatase CD45 leads to th

kinases phosphorylate ITAMs within thecircles). ITAM phosphorylation then lead

activation of the kinase Zap70 and the s

multiple other signalling molecules. One cru

the enzyme phospholipase C -1, leadingmain signalling pathways that drive T-cellkinase Pathway, the protein kinase C

pathway. These pathways culminate in

transcription factors NF-KB, NFAT antranscription and resulting in the differen

effector activity of T-cells.

ease

d begin

R

similar

chain is

. 70) and

and

are then

e CD3

ty that is

gested

generate

[24]. The

V, D, J

e, pre-

andpecificity

l-defined

.

anged

ptor CD3

nd CD8

e (DP)

s

ithin thethymic

C

receptors

mplexes

ctive for

. The

the

. During

cs of

ymocytes to differentiate to become CD4+ or C

en move to the thymic medulla, where they res

mus and enter peripheral circulation [53, 54]. T

t where developing SP thymocytes are extensiv

peripheral tissue-specific antigens [53]. Medull

express thousands of tissue-restricted antigens

complex (TCR and

r CD8 are associated with

e TCR binds to an MHC

tors CD4/CD8 also binddephosphrylation of Src

eir activation, and the Src

receptor complex (redto the recruitment and

ubsequent recruitment of

cial event is activation of

to activation of the threeactivation, the Ras-MAPathway and calcineurin

the activation of the

AP-1, initiating genetiation, proliferation and

D8+ single-

ide for 4-5 days

e thymic

ely screened for

ary thymic

(TRAs) [55], a


23/27

23

phenomenon unique to mTECs known as promiscuous gene expression [56]. Promiscuous gene534

expression allows mTECs to present antigens present in all organs within their MHC molecules535[56]. In addition to mTECs, the thymic medulla also contains a heterogenous population of thymic536

dendritic cells [57]. Two thirds of thymic dendritic cells originate from within the thymus537

(intrathymic DCs) [58], the remaining third are immigrants from peripheral sites (migratory DCs)538

[59]. Intrathymic DCs represent a second route to present mTEC-derived antigents to SP539

thymocytes, which includes secreted molecules, as well as all major subcellular compartments540 including membrane, cytoplasm and nuclear antigens [60]. Migratory DCs are thought to present541

both peripheral self antigens as well as potentially presenting innocuous foreign antigens such as542

those present in gut flora [61]. Thus promiscuous gene expression by mTECs, combined with543antigen presentation by intrathymic and migratory DCs, provides a mirror of self-antigens present544

throughout the organism. SP thymocytes use their time in the thymic medulla to extensively scan545the self antigen-MHC complexes of tMECs, intrathymic DCs and migratory DCs [61]. If a cell has546

too high of an affinity for a self-MHC presented in the thymic medulla, it will trigger activation of547

the TCR, which during this T-cell developmental stage leads to apoptosis [19]. Apoptosis of self-548reactive thymocytes in the thymic medulla efficiently removes auto-reactive thymocytes from the549

T-cell repertoire, a process known as negative selection.550

Supplementary Box 3: Summary of TCR activation and downstream signaling.551

Figure S2 summarizes the primary steps of TCR activation most relevant to our argument. For the552purposes of this review, we have not discussed many of the other receptor ligand interactions553

required for T cell activation (reviewed in [62]). However there are several factors that can affect554

the overall state of T cell activation. With each nave T cell expressing a unique TCR, T cells must555constantly sample MHC molecules if they are to discover the presence of foreign peptide with556

enough affinity to clear the threshold for activation. On the surface, this would seem like a very557

rare occurrence. Indeed, experiments using well characterized pathogens with known558

immunodominant peptides have elegantly shown that the estimated number of cells from a normally559 derived repertoire, which has sufficient affinity to respond to a particular foreign peptide, is560

anywhere from 3x10-4 to 6 x10-7 depending on the pathogen [63, 64]. One factor that contributes to561

appropriate T cell activation in the presence of rare foreign antigens is the length of time the T cell562remains engaged with the APC. Studies show that T cell surveying APC loaded with only563

endogenous peptide display short transient interactions with the APC [65]. When peptides specific564

for the TCR are added, the T cell remains engaged to the APC on the order of hours, which565ultimately results in T cell activation [66, 67]. It is also important to note that other factors can566

affect how the TCR interacts with the peptide/MHC molecule such as TCR clustering, endocytosis567

and degradation of the TCR following engagement (reviewed in [62, 65, 68].568

Supplementary Box 4: A Summary of CGVs role in T-cell activation569

In the context of the adaptive immune system, the hide and environment-dependent release of570

heritable trait differences (CGV) is facilitated by the stochastic recombination of TCR gene571

segments (source of genetic variation), and the stable positive/negative selection conditions in the572thymus. The genome:proteome landscape of T cell receptors contains many distinct proteins that are573

functionally equivalent for the selection conditions of the thymus, i.e. the landscape contains a high574

degree of functional neutrality. T cell maturation statistics suggest that this functional neutrality is575


24/27

24

present in roughly 1-5% of the total TCR G:P landscape. Because selection conditions in the576

thymus cannot distinguish between these neutral variants, this allows for the accumulation of a577genetically diverse repertoire of receptors across the nave T cell population. However, these T cell578

receptors retain structural (entropic and thermodynamic) differences that, while hidden in the initial579

selection context of the thymus, can lead to the expression of highly distinct cell responses (e.g.580

activation) when T-cells are removed from their initial developmental environment and presented581

with antigens.582

In summary, the stochastic recombination of TCR gene segments, and the positive/negative583

selection in the thymus combine to establish non-trivial T cell population properties that result in: 1)584

heritable differences in T-cell receptors that map to the same affinity response for host cells (silent585genetic diversity) 2) the exposure of unique affinities for T cells that are presented with foreign586

antigens (exposure of heritable phenotypic differences), 3) the ability of some of these environment-587

dependent responses to invoke T-cell activation and subsequent clonal expansion (the discovery of588

new selected phenotypes) when presented with foreign antigens. Together, these features establish a589

well-defined role for cryptic genetic variation in an evolving cell population.590

591


25/27

25

Supplementary References592

1. Le Rouzic A., Carlborg . 2008 Evolutionary potential of hidden genetic variation. Trends in Ecology &593Evolution23(1), 33-37.5942. Gibson G., Dworkin I. 2004 Uncovering cryptic genetic variation.Nature Reviews Genetics5(9), 681-690.5953. Schlichting C.D. 2008 Hidden Reaction Norms, Cryptic Genetic Variation, and Evolvability.Annals of the596New York Academy of Sciences1133(1 The Year in Evolutionary Biology 2008), 187-203.5974. McGuigan K., Sgr C.M. 2009 Evolutionary consequences of cryptic genetic variation. Trends in Ecology &598Evolution

24

(6), 305-311.599 5. Kinnison M.T., Hendry A.P. 2001 The pace of modern life II: from rates of contemporary microevolution to600pattern and process. Genetica112(1), 145-164.6016. Hendry A.P., Kinnison M.T. 1999 Perspective: the pace of modern life: measuring rates of contemporary602microevolution.Evolution53(6), 1637-1653.6037. Reznick D.N., Ghalambor C.K. 2001 The population ecology of contemporary adaptations: what empirical604studies reveal about the conditions that promote adaptive evolution. Genetica112(1), 183-198.6058. Barrett R.D.H., Schluter D. 2008 Adaptation from standing genetic variation. Trends in Ecology & Evolution60623(1), 38-44.6079. Hayden E.J., Ferrada E., Wagner A. 2011 Cryptic genetic variation promotes rapid evolutionary adaptation in608an RNA enzyme.Nature474, 92-97.60910. McGuigan K., Nishimura N., Currey M., Hurwit D., Cresko W.A. 2011 Cryptic Genetic Variation and Body610Size Evolution in Threespine Stickleback.Evolution65(4), 1203-1211.61111. Palmer A. 2004 Symmetry breaking and the evolution of development. Science306(5697), 828.61212. Raff M.C. 1973 T and B lymphocytes and immune responses.Nature242(5392), 19.61313. Sallusto F., Geginat J., Lanzavecchia A. 2004 Central memory and effector memory T cell subsets: function,614generation, and maintenance.Annu Rev Immunol22, 745-763.61514. Paul W.E., Benacerraf B. 1977 Functional specificity of thymus-dependent lymphocytes. Science195(4284),6161293.61715. North R.J. 1969 Cellular kinetics associated with the development of acquired cellular resistance. The Journal618of experimental medicine130(2), 299.61916. Rajewsky K. 1996 Clonal selection and learning in the antibody system.62017. Lin J., Harding A., Giurisato E., Shaw A.S. 2009 KSR1 modulates the sensitivity of mitogen-activated protein621kinase pathway activation in T cells without altering fundamental system outputs.Mol Cell Biol29(8), 2082-2091.622(doi:MCB.01634-08 [pii]62310.1128/MCB.01634-08).62418. Altan-Bonnet G., Germain R.N. 2005 Modeling T cell antigen discrimination based on feedback control of625digital ERK responses. PLoS Biol3(11), e356. (doi:05-PLBI-RA-0208R2 [pii]62610.1371/journal.pbio.0030356).62719. Daniels M.A., Teixeiro E., Gill J., Hausmann B., Roubaty D., Holmberg K., Werlen G., Hollander G.A.,628Gascoigne N.R., Palmer E. 2006 Thymic selection threshold defined by compartmentalization of Ras/MAPK signalling.629Nature444(7120), 724-729. (doi:nature05269 [pii]63010.1038/nature05269).63120. Prasad A., Zikherman J., Das J., Roose J.P., Weiss A., Chakraborty A.K. 2009 Origin of the sharp boundary632that discriminates positive and negative selection of thymocytes. Proc Natl Acad Sci U S A106(2), 528-533.633(doi:0805981105 [pii]63410.1073/pnas.0805981105).63521. Das J., Ho M., Zikherman J., Govern C., Yang M., Weiss A., Chakraborty A.K., Roose J.P. 2009 Digital636signaling and hysteresis characterize ras activation in lymphoid cells. Cell136(2), 337-351. (doi:S0092-8674(08)01631-6370 [pii]63810.1016/j.cell.2008.11.051).63922. Chakraborty A.K., Das J., Zikherman J., Yang M., Govern C.C., Ho M., Weiss A., Roose J. 2009 Molecular640origin and functional consequences of digital signaling and hysteresis during Ras activation in lymphocytes. Sci Signal6412(66), pt2. (doi:scisignal.266pt2 [pii]64210.1126/scisignal.266pt2).64323. Schatz D. 2004 V (d) j recombination.Immunological Reviews200(1), 5-11.64424. Davis M.M., Bjorkman P.J. 1988 T-cell antigen receptor genes and T-cell recognition.64525. Starr T., Jameson S., Hogquist K. 2003 Positive and negative selection of T cells.Annual Review of646Immunology21, 139.64726. Arstila T.P., Casrouge A., Baron V., Even J., Kanellopoulos J., Kourilsky P. 1999 A direct estimate of the648human alphabeta T cell receptor diversity. Science286(5441), 958.64927. Waddington C.H. 1942 Canalization of Development and the Inheritance of Acquired Characters.Nature650150(3811), 563.65128. Waddington C.H. 1957 The strategy of the genes, Allen & Unwin London.652


26/27

26

29. Badyaev A.V. 2005 Stress-induced variation in evolution: from behavioural plasticity to genetic assimilation.653Proceedings of the Royal Society B: Biological Sciences272(1566), 877.65430. Whitacre J.M. (In press) Genetic and environment-induced pathways to innovation: on the possibility of a655universal relationship between robustness and adaptation in complex biological systems656(http://www.springerlink.com/content/0577136586542281/)Evol Ecol. (doi:10.1007/s10682-011-9464-z).65731. Waddington C.H. 1953 Genetic Assimilation of an Acquired Character.Evolution7(2), 118-126.65832. Waddington C.H. 1957 The Strategy of the Genes: A Discussion of Some Aspects of Theoretical Biology, Allen659& Unwin.66033. Schmalhausen I., Dobzhansky T. 1949 Factors of evolution: the theory of stabilizing selection, Blakiston Co.661Philadelphia.66234. West-Eberhard M. 2005 Developmental plasticity and the origin of species differences. Proceedings of the663National Academy of Sciences102(Suppl 1), 6543.66435. West-Eberhard M. 2003Developmental plasticity and evolution, Oxford University Press, USA.66536. Pigliucci M., Murren C., Schlichting C. 2006 Phenotypic plasticity and evolution by genetic assimilation.J666Exp Biol209(12), 2362.66737. Tian T., Olson S., Whitacre J.M., Harding A. 2011 The origins of cancer robustness and evolvability.668Integrative Biology3, 17-30. (doi:10.1039/c0ib00046a).66938. Gerlinger M., Swanton C. 2010 How Darwinian models inform therapeutic failure initiated by clonal670heterogeneity in cancer medicine.Br J Cancer103(8), 1139-1143.67139. Read A.F., Day T., Huijben S. 2011 The evolution of drug resistance and the curious orthodoxy of aggressive672chemotherapy. Proceedings of the National Academy of Sciences108(Supplement 2), 10871.67340. Frank S. 2007 Maladaptation and the paradox of robustness in evolution. PLoS One2(10), 1021.67441. Wagner A. 2008 Robustness and evolvability: a paradox resolved. Proceedings of the Royal Society of London,675Series B: Biological Sciences275, 91-100.67642. Ciliberti S., Martin O.C., Wagner A. 2007 Innovation and robustness in complex regulatory gene networks.677Proceedings of the National Academy of Sciences, USA104(34), 13591-13596.67843. Whitacre J.M., Bender A. 2010 Degeneracy: a design principle for achieving robustness and evolvability.679Journal of Theoretical Biology263(1), 143-153. (doi:S0022-5193(09)00534-7 [pii]68010.1016/j.jtbi.2009.11.008).68144. Whitacre J.M. (In press) Genetic and environment-induced pathways to innovation: on the possibility of a682universal relationship between robustness and adaptation in complex biological systems683(http://www.springerlink.com/content/0577136586542281/)Evolutionary Ecology. (doi:10.1007/s10682-011-9464-z).68445. Whitacre J.M., Atamas S.P. (in press) The Diversity Paradox: Resolving Darwins Unspoken Dilemma.685Biology Direct.68646. Le Douarin N.M., Jotereau F.V. 1975 Tracing of cells of the avian thymus through embryonic life in687interspecific chimeras.J Exp Med142(1), 17-40.68847. Nitta T., Murata S., Ueno T., Tanaka K., Takahama Y. 2008 Thymic microenvironments for T-cell repertoire689

formation.Adv Immunol99, 59-94. (doi:S0065-2776(08)00603-2 [pii]690 10.1016/S0065-2776(08)00603-2).69148. Petrie H.T., Hugo P., Scollay R., Shortman K. 1990 Lineage relationships and developmental kinetics of692immature thymocytes: CD3, CD4, and CD8 acquisition in vivo and in vitro.J Exp Med172(6), 1583-1588.69349. Petrie H.T., Pearse M., Scollay R., Shortman K. 1990 Development of immature thymocytes: initiation of694CD3, CD4, and CD8 acquisition parallels down-regulation of the interleukin 2 receptor alpha chain.Eur J Immunol69520(12), 2813-2815. (doi:10.1002/eji.1830201243).69650. Witt C.M., Raychaudhuri S., Schaefer B., Chakraborty A.K., Robey E.A. 2005 Directed migration of697positively selected thymocytes visualized in real time. PLoS Biol3(6), e160. (doi:04-PLBI-RA-0989R2 [pii]69810.1371/journal.pbio.0030160).69951. Ashton-Rickardt P.G., Van Kaer L., Schumacher T.N., Ploegh H.L., Tonegawa S. 1993 Peptide contributes to700the specificity of positive selection of CD8+ T cells in the thymus. Cell73(5), 1041-1049. (doi:0092-8674(93)90281-T701[pii]).70252. Sebzda E., Wallace V.A., Mayer J., Yeung R.S., Mak T.W., Ohashi P.S. 1994 Positive and negative thymocyte703

selection induced by different concentrations of a single peptide. Science263(5153), 1615-1618.704 53. Takahama Y. 2006 Journey through the thymus: stromal guides for T-cell development and selection.Nat Rev705Immunol6(2), 127-135. (doi:nri1781 [pii]70610.1038/nri1781).70754. McCaughtry T.M., Wilken M.S., Hogquist K.A. 2007 Thymic emigration revisited.J Exp Med204(11), 2513-7082520. (doi:jem.20070601 [pii]70910.1084/jem.20070601).71055. Anderson M.S., Venanzi E.S., Klein L., Chen Z., Berzins S.P., Turley S.J., von Boehmer H., Bronson R.,711Dierich A., Benoist C., et al. 2002 Projection of an immunological self shadow within the thymus by the aire protein.712Science298(5597), 1395-1401. (doi:10.1126/science.10759587131075958 [pii]).714


27/27

56. Derbinski J., Schulte A., Kyewski B., Klein L. 2001 Promiscuous gene expression in medullary thymic715epithelial cells mirrors the peripheral self.Nat Immunol2(11), 1032-1039. (doi:10.1038/ni723716ni723 [pii]).71757. Wu L., Shortman K. 2005 Heterogeneity of thymic dendritic cells. Semin Immunol17(4), 304-312.718(doi:S1044-5323(05)00039-4 [pii]71910.1016/j.smim.2005.05.001).72058. Donskoy E., Goldschneider I. 2003 Two developmentally distinct populations of dendritic cells inhabit the721adult mouse thymus: demonstration by differential importation of hematogenous precursors under steady state722conditions.J Immunol170(7), 3514-3521.72359. Li J., Park J., Foss D., Goldschneider I. 2009 Thymus-homing peripheral dendritic cells constitute two of the724three major subsets of dendritic cells in the steady-state thymus.J Exp Med206(3), 607-622. (doi:jem.20082232 [pii]72510.1084/jem.20082232).72660. Koble C., Kyewski B. 2009 The thymic medulla: a unique microenvironment for intercellular self-antigen727transfer.J Exp Med206(7), 1505-1513. (doi:jem.20082449 [pii]72810.1084/jem.20082449).72961. Klein L., Hinterberger M., Wirnsberger G., Kyewski B. 2009 Antigen presentation in the thymus for positive730selection and central tolerance induction.Nat Rev Immunol9(12), 833-844. (doi:nri2669 [pii]73110.1038/nri2669).73262. Rudolph M.G., Stanfield R.L., Wilson I.A. 2006 How TCRs bind MHCs, peptides, and coreceptors.Annu Rev733Immunol24, 419-466. (doi:10.1146/annurev.immunol.23.021704.115658).73463. Moon J.J., Chu H.H., Pepper M., McSorley S.J., Jameson S.C., Kedl R.M., Jenkins M.K. 2007 Naive CD4(+)735T cell frequency varies for different epitopes and predicts repertoire diversity and response magnitude.Immunity27(2),736203-213. (doi:S1074-7613(07)00366-4 [pii]73710.1016/j.immuni.2007.07.007).73864. Alanio C., Lemaitre F., Law H.K., Hasan M., Albert M.L. 2010 Enumeration of human antigen-specific naive739CD8+ T cells reveals conserved precursor frequencies.Blood115(18), 3718-3725. (doi:blood-2009-10-251124 [pii]74010.1182/blood-2009-10-251124).74165. Salazar-Fontana L.I., Bierer B.E. 2001 T-lymphocyte coactivator molecules. Curr Opin Hematol8(1), 5-11.74266. Miller M.J., Wei S.H., Parker I., Cahalan M.D. 2002 Two-photon imaging of lymphocyte motility and antigen743response in intact lymph node. Science296(5574), 1869-1873. (doi:10.1126/science.10700517441070051 [pii]).74567. Miller M.J., Safrina O., Parker I., Cahalan M.D. 2004 Imaging the single cell dynamics of CD4+ T cell746activation by dendritic cells in lymph nodes.J Exp Med200(7), 847-856. (doi:jem.20041236 [pii]74710.1084/jem.20041236).74868. Smith-Garvin J.E., Koretzky G.A., Jordan M.S. 2009 T cell activation.Annu Rev Immunol27, 591-619.749(doi:10.1146/annurev.immunol.021908.13270675010.1146/annurev.immunol.021908.132706 [pii]).751

752753

754

Documents

T cell adaptive immunity proceeds through environment-induced adaptation from the exposure of cryptic genetic variation