Mystery Rays from Outer Space

As the overwhelming majority of the mutations in cancer cells are unrelated to malignancy, the mutation-generated epitopes shall be specific for each individual tumor, and constitute the antigenic fingerprint of each tumor. These calculations highlight the benefits for personalization of immunotherapy of human cancer, and in view of the substantial pre-existing antigenic repertoire of tumors, emphasize the enormous potential of therapies that modulate the anti-cancer immune response by liberating it from inhibitory influences.

–Srivastava N, Srivastava PK (2009)
Modeling the Repertoire of True Tumor-Specific MHC I Epitopes in a Human Tumor.
PLoS ONE 4(7): e6094. doi:10.1371/journal.pone.0006094

(My emphasis)

(See many of my previous posts here for more information)

Dogs & Wolf (Gotthilf Heinrich von Schubert, 1872) Naturgeschichte der Säugethiere: mit colorirten Abbildungen zum Anschauugs-Unterricht für die Jugend. (Esslingen : Schreiber, 1872)

Cancer is a creepy disease. Your own cells turn on you, mindlessly and blindly destroying themselves — because the only way a cancer can survive is for its host to survive; unlike viruses, cancers don’t spread from their original carrier to new hosts. Each cancer is a new and unique event, and each cancer is a terminal event, with no future, no children, no transmission. ¹ Cancers don’t spread between individuals, because they would essentially be tissue transplants; if you try to transplant skin (or whatever) between two animals, the skin will be rapidly rejected because of mismatched MHC alleles, and the same is true for tumors.

In theory, tumors could spread from one identical twin to another (I don’t know of cases where this has occurred, though). Tumors might also theoretically spread within a population that has little MHC diversity. Cheetahs, for example, notoriously don’t have much MHC diversity,² probably due to a genetic bottleneck some 10,000 years ago,³ and tolerate random skin grafts very well;⁴ it’s conceivable that a tumor that arose in one cheetah could spread throughout the population. But MHC is generally diverse, and becomes diverse very rapidly, so these sort of low-diversity populations are unusual (though they may be more common than generally realized). And so, as I say, tumors are dead ends and don’t spread from their original host.

But there are always exceptions, and we know of a couple exceptions for this principle as well. Tasmanian Devil Facial Tumor is one; canine transmissible venereal tumor (CTVT) is the other. The last time I talked about this, I said that there wasn’t anything particularly surprising about TDFT. I’ve changed my mind about that, for reasons I’ll talk about later. Here I want to give a quick update on CTVT, which is still even more interesting than TDFT.

CTVT is just what it sounds like: an infectious tumor of dogs. It can be transmitted from one dog to the next, mainly during sex. There’s no sign of a viral agent, and in fact the tumor cells are distinct from the host — that is, the tumor cells are not host cells that have been transformed by, say, a virus, but are foreign agents altogether, cells from an ancient animal that have been passed along in this way, generation after generation, spreading throughout the world. Cells from a CTVT on a German Shepherd in Italy are virtually identical to those from a CTVT on a poodle in Brazil, but the tumor’s cells are very different from their hosts’: “… the differences in genome content between dog and CTVT are substantially greater than those between different CTVT samples.“ ⁵

The last time I talked about this I cited a paper⁶ that showed that the original host of CTVT was actually very ancient: “Our analysis of divergence of microsatellites indicates that the tumor arose between 200 and 2500 years ago.” ⁶ Amazingly enough,  it seemed that the tumor had been passed between dogs perhaps for a couple of thousand years.

But it turns out that it’s even more amazing than that: The 2500 year estimate turns out to be an underestimate. “The estimated time of origin of CTVT is 6500–65,000 years ago. … The cycle of infection currently takes about six months; if this has been true for, say, 15,000 years, then there have been 30,000 transmission events.”⁵ That spans the age where dogs were domesticated and diverged from wolves , about 10,000-15,000 years ago; and in fact, based on genome analysis of the tumors, the tumors did arise in wolves, not dogs. ⁷

This raises a fascinating scenario (which I didn’t see spelled out in Reddeck et al, though I think they were hinting at it). It’s still mysterious how CTVT spreads between hosts. As I said, tumors should act like tissue grafts, which are very rapidly rejected if they’re not MHC matched (as is the case with the CTVT and modern dogs). But tumors could, theoretically, spread within a population that has minimal MHC diversity — that is, an inbred population, that has recently undergone a genetic bottleneck. Such a bottleneck very likely happened during the domestication of dogs, as they evolved from wolves. Was the original CTVT a tumor of one of the very first domesticated proto-dogs — an inbred proto-dog,⁸ a member of a small and inbred population — a spontaneous, fairly ordinary tumor that then spread to the first host’s MHC-matched neighbours in the small and inbred population? That would not be so mysterious; we know tumors can be transplanted between genetically identical hosts, that’s done with lab mice all the time. Then, as the domestic dog population expanded and spread over the world, the tumor would expand and spread along with it, and would have the opportunity to undergo natural selection, adapting to the newly-diverse MHC and to the immune responses that arose after the tumor did. We don’t know, in molecular terms, how the tumor has adapted — we don’t know why it’s able to spread in grow in spite of host immune responses. But at least we can now imagine how the tumor could have reached this point. Thirty thousand transmission events — a couple of million cell divisions — is a long time for a cancer to adapt.

1. Even most viral cancers are probably dead ends, because many viral cancers represent abortive viral replication.[↩]
2. DNA variation of the mammalian major histocompatibility complex reflects genomic diversity and population history.
   Yuhki N, O’Brien SJ
   Proc Natl Acad Sci USA (1990) 87:836–840.[↩]
3. Molecular Genetic Insights on Cheetah (Acinonyx jubatus) Ecology and Conservation in Namibia
   Laurie L. Marker, Alison J. Pearks Wilkerson, Ronald J. Sarno, Janice Martenson, Christian Breitenmoser-Würsten, Stephen J. O’Brien, and Warren E. Johnson
   Journal of Heredity 2008 99(1):2-13; doi:10.1093/jhered/esm081[↩]
4. Genetic basis for species vulnerability in the cheetah.
   O’Brien SJ, Roelke ME, Marker L, Newman A, Winkler CA, Meltzer D, Colly L, Evermann JF, Bush M, Wildt DE.
   Science. 1985 Mar 22;227(4693):1428-34

   It was suggested at the time, and is widely believed, that cheetahs are particularly susceptible to infectious diseases for this reason, but I think the evidence for this is very weak. Cheetahs have high mortality when infected with feline infectious peritonitis virus, but it’s not unusual for viruses infecting new species to have a high mortality — think SARS and avian influenza in hunans, canine distemper in seals, and feline (or something) parvovirus in dogs, all of which had high mortality in their new, MHC-diverse, host[↩]

5. Rebbeck, C., Thomas, R., Breen, M., Leroi, A., & Burt, A. (2009). ORIGINS AND EVOLUTION OF A TRANSMISSIBLE CANCER Evolution DOI: 10.1111/j.1558-5646.2009.00724.x[↩][↩]
6. Murgia, C., Pritchard, J. K., Kim, S. Y., Fassati, A., and Weiss, R. A. (2006). Clonal origin and evolution of a transmissible cancer. Cell 126, 477-487 .[↩][↩]
7. To be fair to Murgia, they pointed out that their 250-2500 year estimate was for the divergence of the tumor, not its origin, and they specifically raised the possibility that the tumor originated earlier than dogs.[↩]
8. Reddeck et al. point out that the tumor did, in fact, arise from an inbred animal[↩]

You can go to the most prestigious medical center in the world and ask “How is my immune system?” and, after a short period of eye rolling and looks of amused incomprehension, you might (if they don’t just throw you out) be offered a white blood cell count (which you should probably decline). … How did we arrive at this state of affairs? A good case can be made that the mouse has been so successful at uncovering basic immunologic mechanisms that now many immunologists rely on it to answer every question. … Well, except that mice are lousy models for clinical studies. This is readily apparent in autoimmunity (von Herrath and Nepom, 2005) and in cancer immunotherapy (Ostrand-Rosenberg, 2004), where of dozens (if not hundreds) of protocols that work well in mice, very few have been successful in humans. ¹

Cytotoxic T lymphocyte attacking a tumor cell

The above (emphasis added) is from a recent manifesto from immunology giant Mark Davis.¹ (He isn’t the first to make this point, of course, but when Mark Davis speaks, immunologists listen.)  Davis’s suggested solution was, among other things, to start using humans as their own models, taking advantage of the large numbers of humans who are routinely screened and overcoming the lack of experimental control by taking a “systems” approach, including large-scale data collection from healthy and ill people and large-scale informatics as part of the analysis.² (He also comments on the “humanized” mouse approach that I mentioned briefly the other day.)

Here’s an example of the power of human models, though it’s not exactly what Davis is describing.³  I’ve mentioned before the evidence that cancers in mice are controlled by the immune system (for example, here and links therein).  In those experiments, mutant mice, lacking one or more components of the immune response, were shown to be predisposed to cancer.  There are also a couple of human studies that indicate the same thing; people on long-term immunosuppression (as in transplant recipients) are somewhat more likely to get certain kinds of cancer, for example.

CTL attacking a tumor cell

One advantage of using humans as models of their own diseases is that there are an amazing number of well-documented mutations and disease-associated genes in humans.  Human disease is taken very seriously, it’s well funded (at least in comparison to, say, dog and cat disease); even diseases that are very rare can be identified in humans and the gene variant identified.  One such disease is Type II familial hemophagocytic lymphohistiocytosis (FHL), a rare, rapidly fatal disease caused by mutations in the perforin gene. ⁴  Perforin is important in cytotoxic T lymphocyte function, and it’s one of the immune genes that has been shown to be important in preventing cancers in mice.⁵

Are patients with Type II FHL at risk of developing cancer, like mice with targeted perforin mutations?  It’s not an easy question to ask, because most patients die relatively young.  But because these are humans, this very rare disease is nevertheless well documented, and the authors were able to search the literature for a suitable subset of patients:

… we identified a subgroup of individuals from nonconsanguineous families who possessed 2 mutated PRF1 alleles but whose onset of FHL was markedly delayed (the age at onset of 10 years or older) or even abolished. A total of only 23 such cases could be identified in the entire literature …  Ten of the individuals (Patients 14–23 inTable 1) developed manifestations of FHL without any other significant infectious or neoplastic sequelae reported. … Remarkably, in 11 of these 13 individuals (or 48% of the entire cohort of 23), the primary clinical presentation was with either B or T cell lymphoma or acute or chronic leukemia of lymphoid origin. … The very high frequency of hematological cancers in this 23-patient cohort …  is vastly in excess of that in the general population.³

There’s a lot of other interesting stuff in the paper, but this is enough to make the point: Just as in mice, perforin in humans (and therefore, the immune system) is important in preventing cancer.

I’ll leave with this now-familiar observation from the paper:

It is clearly problematic to extrapolate experimental data from inbred mouse strains to an outbred human setting where such evidence is far more difficult to gather.³

1. DAVIS, M. (2008). A Prescription for Human Immunology Immunity, 29 (6), 835-838 DOI: 10.1016/j.immuni.2008.12.003[↩][↩]
2. Again, of course, Davis isn’t the first to advocate this approach.[↩]
3. Chia, J., Yeo, K., Whisstock, J., Dunstone, M., Trapani, J., & Voskoboinik, I. (2009). Temperature sensitivity of human perforin mutants unmasks subtotal loss of cytotoxicity, delayed FHL, and a predisposition to cancer Proceedings of the National Academy of Sciences, 106 (24), 9809-9814 DOI: 10.1073/pnas.0903815106[↩][↩][↩]
4. Familial hemophagocytic lymphohistiocytosis. Primary hemophagocytic lymphohistiocytosis.
   Henter JI, Aricò M, Elinder G, Imashuku S, Janka G.
   Hematol Oncol Clin North Am. 1998 Apr;12(2):417-33[↩]
5. Perforin-mediated Cytotoxicity Is Critical for Surveillance of Spontaneous Lymphoma.
   Mark J. Smyth, Kevin Y.T. Thia, Shayna E.A. Street, Duncan MacGregor, Dale I. Godfrey, and Joseph A. Trapani.
   The Journal of Experimental Medicine, Volume 192, Number 5, September 5, 2000 755-760[↩]

Caco2 colon carcinoma cells

One of my long-standing questions now has at least a partial answer, or maybe a pathway toward a partial answer.

My question was, “Why are different tumors the same?” That is, why do tumors of the same type often seem to have similar immunological changes?

Viruses of the same kind (all herpes simplex viruses, say) all avoid immunity in the same way because they all have a common ancestor from which they inherited their immune evasion molecules. But tumors have no common ancestor; each tumor has to grope around and find its own solution to common problems independently. So why should tumors of a particular tissue converge on common solutions? What would make a certain pathway a simple solution for colon carcinomas, but not for bladder tumors? And yet, apparently that’s what happens. Even though many different tumor types become non-immunogenic, tumors of a particular tissue type often reach non-immunogenicity by a similar path. For example:

… distinct molecular events underlie HLA class I loss, depending on the aetiology of the tumours; Lynch syndrome-related cancers presented with mutations in the β₂-m molecule, while sporadic microsatellite-unstable tumours mainly showed alterations in the antigen-processing machinery components ¹

When I posed this question a couple of months ago, I didn’t have an answer. I said

Part of the answer may be that the particular oncogenes associated with different tumor types lead to particular transcriptional hot-spots, and being a transcriptional hot-spot makes the region a mutational hot-spot as well, but at least as I understand it that’s not enough to account for the trends.

EH, in the comments, suggested a couple more possibilities: Perhaps destruction of a particular pathway is “a nice side benefit of destroying some yet undiscovered tumor suppressor important for melanoma or colon cancer? Or maybe the loss of certain repair genes common to a cancer type (like BRCA) leads to selective chromosomal instability in parts of chromosome 6?“

It turns out, unsurprisingly, that I’m not the only one to ask the question. There is no general answer, but apparently a partial answer has been kicking around for a while now, and Hans Morreau’s group is starting to put some of the pieces together.² As their model, Morreau’s group looked at MUTYH-associated polyposis (“MAP”)-associated tumors. (MAP is a heritable defect in DNA repair, so MAP patients have high rates of mutagenesis and usually develop colon carcinomas.) They reasoned that

MAP tumours could be more prone to stimulate a cytotoxic T-cell-mediated immune response, due to their frequent generation of aberrant peptides. Hence, these tumours could also be subjected to a strong selective pressure favouring the outgrowth of cancer cells that acquire an immune evasive phenotype.

DNA Repair

In other words, the argument is that the group of tumors that lose (or reduce) DNA repair are much more likely to throw out mutant proteins. These mutant proteins are targets for the immune response (because they’re no longer “self” antigens), so to survive the immune attack the tumor has a strong selection for loss of immunogenicity (and also has the high mutation rate that allows them to rapidly mutate away from immunogenicity).

Sure enough, the tumors did frequently (72%) have defects in antigen presentation. (In fact, because of the way they measured HLA expression — by immunohistochemistry rather than sequencing — I would bet that the rate of functional defects was actually much higher than that.) They conclude that this “provides additional evidence that tumours carrying defects in DNA base repair mechanisms are more prone to undergo immune escape mechanisms.” ²  Since they don’t formally compare to other tumor types, I don’t think they can really say “more prone” — you’d have to use the same techniques to look at tumors from non-MAP patients to be able to say that. Still, I do think that is a significantly higher rate than has been turned up in previous studies using similar techniques,³ so I’ll tentatively accept that conclusion.

This doesn’t really explain why similar tumors target similar components, but it’s at least a conceptual connection between different tumor types and an underlying pathway. I’d be interested in a more large-scale screen, looking at cancers with stronger and weaker mutator phenotypes to see if common pathways emerge. One may already have popped up, since the authors note here that expression of β₂-m was frequently lost⁴ in several tumors with DNA repair defects:

Although speculative, it is interesting to underline that carcinomas derived from both MAP and Lynch syndromes preferentially lose β₂-m expression coupled to HLA class I deficiencies. A functional explanation for these observations remains elusive, but perhaps distinct reactions (both qualitative and quantitative) by the immune system, depending on the age of onset of the tumours, could condition the type of mechanisms that lead to HLA class I expression deficiencies. ²

Again, we would need to compare to other tumor types to see if this really is more frequent, but overall it feels as if there’s a hint of a pathway here. At least there are some specific questions that can be asked.

1. de Miranda, N., Nielsen, M., Pereira, D., van Puijenbroek, M., Vasen, H., Hes, F., van Wezel, T., & Morreau, H. (2009). MUTYH-associated polyposis carcinomas frequently lose HLA class I expression-a common event amongst DNA-repair-deficient colorectal cancers The Journal of Pathology DOI: 10.1002/path.2569
   Referencing
   Dierssen JWF, de Miranda NFCC, Ferrone S, van Puijenbroek M, Cornelisse CJ, Fleuren GJ, et al. HNPCC versus sporadic microsatellite-unstable colon cancers follow different routes toward loss of HLA class I expression. BMC Cancer 2007; 7: 33[↩]
2. de Miranda, N., Nielsen, M., Pereira, D., van Puijenbroek, M., Vasen, H., Hes, F., van Wezel, T., & Morreau, H. (2009). MUTYH-associated polyposis carcinomas frequently lose HLA class I expression-a common event amongst DNA-repair-deficient colorectal cancers The Journal of Pathology DOI: 10.1002/path.2569[↩][↩][↩]
3. Other studies, especially those of Soldano Ferrone, have turned up much higher rates of functional HLA class I defects, but they’ve used more focused, and more difficult and expensive, techniques to reach that conclusion.[↩]
4. β₂-m is a physical component of the MHC class I complex[↩]

We know that the immune system can destroy tumors. We also know, unfortunately, that by the time we see a tumor, immunity probably won’t destroy the tumor. There are lots of reasons for that. One is that tumors are essentially part of the normal body, so it’s normal for the immune system to ignore them. It looks as if you need to have immunity that’s just right to get rid of a tumor.

Tumors arise from normal self cells,¹ that the immune response has been programmed to ignore. Now, the process of becoming a tumor is not normal, and so tumors are not entirely normal self any more — meaning that there are likely to be some targets in most if not all tumors. But in all but the most reckless tumors the differences between abnormal and normal are relatively small, compared to, say, a virus-infected cell that contains many potential targets.

There’s actually a long list of known tumor antigens; the T-cell tumor peptide database lists many hundreds of them. But most are not truly specific for the tumor.  The’re actually normal self antigens; they’re derived from proteins that are overexpressed in tumors, or that are differentiation antigens or “cancer-germline” antigens that are normally also found in self tissues. What’s more, these normal self antigens are the most interesting tumor antigens, as far as clinical utility is concerned. Mutations can make brand-new, non-self targets for the immune system, but they’re going to be sporadic targets, often unique to individual tumors — not something you can prepare for. The normal antigens, though, are likely to be predictable, common targets; it’s conceivable that tumor vaccines can be prepared in advance.

Human melanoma cell

If these antigens were common (which they are, in some tumor types — like melanoma), and they were good targets for the immune system, then we wouldn’t see much cancer. We do see melanomas quite often, and part of the reason may be that the immune system generally responds quite weakly to these antigens.  Why is that? And, more to the point, how can we make the immune system respond more strongly? A recent paper in the Journal of Experimental Medicine² offers answers for both of these questions.

From work in the past couple of years, we now have decent estimates of how many T cells there are that can react with any particular target. (See here and here for my discussion of the earlier papers.) A reasonably strong immune response to a non-self epitope might originate from maybe 100 or so precursor T cells. There’s a rather wide range of frequency for these precursor cells, say from 20 to 1000; and to some extent, the fewer T cells there are the weaker (the less immunodominant) the immune response.

We expect T cells against normal self targets to be less common, because they should be eliminated as they mature in the thymus. Some may survive, though, and we would count on these survivors to attack the normal (albeit overexpressed, or abnormally present) target in the cancer cells. But just how rare are they?

Rizzuto et al say they’re really rare (this was in mice, by the way); at least ten times less abundant than T cells against non-self antigens.  If you look at the range I gave for “normal” precursors, that could mean there are fewer than 5 or 10 precursors.  If the average is “fewer than five”, then quite possibly some mice have only two, or one, or no precursors.  You can’t have much of a response with no precursors.

So there’s a weak anti-tumor response because there aren’t many T cells in the body that can respond to the normal self targets in the tumor. That’s not really a surprise, but it does raise the question, What if there were more of the T cells? To ask that question, Rizzuto et al. tried transferring more of these precursor T cells into tumor-bearing mice — starting at around the normal level for a precursor to non-self antigen, and going up from there — and then vaccinating with the appropriate target.

The effects were pretty dramatic. With no supplemental T cells (that is, with the natural, very low, level of T cell precursors) the mice all died of the tumor quickly. At the middle of the range, almost all of the mice rejected the tumor. And at the highest levels of transfers? The mice all died again. Having enough T cells to respond was protective, but putting in too many made them useless.

These results identify vaccine-specific CD8+ precursor frequency as a remarkably significant predictor of treatment and side-effect outcome. Paradoxically, above a certain threshold there is an inverse relationship between pmel-1 clonal frequency and vaccine-induced tumor rejection.²

Mouse melanoma cell

(My emphasis) This paradoxical effect is probably because the numerous T cells started to compete with each other so that none of them were properly activated; they only saw effective-looking polyfunctional T cells at the lower transfer levels.

In other words, if you’re going to transfer T cells to try to eliminate a tumor, more is not necessarily better. Quality and quantity are both important factors, and quantity helps determine quality.

One question I have is how this relates to tumor immune evasion. Many tumor types  acquire mutations, as they develop, that block presentation of antigen to T cells. Are these mutations perhaps only partially effective — giving the tumors sufficient protection against the tiny handful of natural precursors they “expect” to deal with, but not against a larger attack after, say, vaccination — or are they more complete, and protective even if the optimal number of T cells are transfered? I’d guess that it would depend on the tumor, but it looks as if it might be a relevant question and it would be nice to have more than a guess.

Our results show that combining lymphodepletion with physiologically relevant numbers of naive tumor-specific CD8+ cells and in vivo administration of an effective vaccine generates a high-quality, antitumor response in mice. This approach requires strikingly low numbers of naive tumor-specific cells, making it a new and truly potent treatment strategy.   ²

1. I’m ignoring here crazy things like the contagious tumors of Tasmanian Devils and dogs[↩]
2. Rizzuto, G., Merghoub, T., Hirschhorn-Cymerman, D., Liu, C., Lesokhin, A., Sahawneh, D., Zhong, H., Panageas, K., Perales, M., Altan-Bonnet, G., Wolchok, J., & Houghton, A. (2009). Self-antigen-specific CD8+ T cell precursor frequency determines the quality of the antitumor immune response Journal of Experimental Medicine, 206 (4), 849-866 DOI: 10.1084/jem.20081382[↩][↩][↩]

Hierarchical clustering of breast carcinomas1

Something that’s puzzled me for years is why the same kinds of tumors tend to have the same kinds of immune evasion mechanisms. And I’m not going to give an answer, just trying to share the confusion a little.

What I mean is this:

It has been demonstrated that human tumors of distinct histology express low or downregulated MHC class I surface antigens … The distinct frequency of MHC class I abnormalities is caused by total HLA class I antigen loss, HLA class I down-regulation as well as loss or down-regulation of HLA class I allo-specificities. However, the frequency and mode of these defects significantly varied between the types of tumors analysed and could be associated in some cases with microsatellite instability. ²

(My emphasis) As I’ve noted here several times (most specifically here) tumors very often evade the immune system as they mature. Cytotoxic T lymphocytes (CTL) can control tumors in the tumors’ eary stages, but by the time we detect a tumor clinically the tumor is almost always resistant to the immune system. They do this in various ways, including inducing regulatory T cells, but also by mutating themselves to make themselves invisible to CTL (and other components of the immune system, but let’s keep it simpler for the moment).

There are a myriad ways for a tumor to become invisible, at the molecular level.  The MHC class I antigen presentation pathway is long and complex, and for any partiuclar tumor there are likely to be many different bottlenecks, points of attack.  Since tumors are all independent events³, so at first, and even second, glance, there’s no obvious reason why tumors of the same type should find a similar approach.  That is, just because two colon carcinomas look the same histologically in two different individuals, there’s no link between them.  ⁴ Why should colon carcinomas avoid CTL using one set of mutations, while, say, breast cancers use a different set of mutations? Yet apparently, that’s what tends to happen; for example:

Mutations or deletions in β2-m were detected in colon carcinoma (21%), melanoma (15%) and other tumors (<5%). So far, no mutations in β2-m have been found in RCC lesions, bladder and laryngeal tumors despite MHC class I loss or downregulation. … haplotype loss was found in head and neck squamous cell carcinoma (HNSCC) with a frequency of 36%, whereas in renal cell carcinoma (RCC) LOH only occurs in approximately 12% of tumor lesions analyzed. ²

If we saw these patterns only with virus-associated cancers, such as cervical carcinomas and even hepatic carcinomas, there would at least be a common link, but these tumors are not (as far as we know) caused by viruses in humans.

Part of the answer may be that the particular oncogenes associated with different tumor types lead to particular transcriptional hot-spots, and being a transcriptional hot-spot makes the region a mutational hot-spot as well, but at least as I understand it that’s not enough to account for the trends.

So why are particular MHC abnormalities linked to tumor type?  Anyone?

1. Turashvili et al. BMC Cancer 2007 7:55   doi:10.1186/1471-2407-7-55[↩]
2. Seliger, B. (2008). Molecular mechanisms of MHC class I abnormalities and APM components in human tumors Cancer Immunology, Immunotherapy, 57 (11), 1719-1726 DOI: 10.1007/s00262-008-0515-4[↩][↩]
3. barring such weird things as canine transmissible venereal tumor and Tasmanian Devil facial tumors; see here for more on those[↩]
4. The comparison is, of course, viruses.  A herpesvirus of chickens, and one of humans, may both use immune evasion mechanisms, but they have a common ancestor even if it’s a couple of hundred million years ago.[↩]

Metchnikov: “Lecons sur la pathologie” (1892)

There are times when you just feel like the universe is out to get you. For example, we know that inflammation can drive tumor formation; but a paper just came out that suggests reducing inflammation can also drive tumor formation. ¹ It doesn’t seem fair.

I’ve previously mentioned the link between inflammation and tumorigenesis, which is probably at least partly because the inflammation produces reactive oxygen and nitrogen species (RONS ) that are tumorigenic.

I’ve also talked about the link between reduced inflammation and ongoing tumors (for example, here, here, and here). What seems to be going on here is that regulatory T cells (TRegs) are induced by tumors, and these TRegs shut down anti-tumor immunity.

So far, these findings aren’t really contradictory. Increasing inflammation before a tumor is present makes tumors more likely to form. After the tumor has formed, reducing inflammation makes the tumor more likely to persist. The universe-is-against-us part comes from the suggestion that reducing inflammation (via TRegs) before tumor formation, also makes the tumors more likely to form.

This may be a special case. The paper from Philip Dennis’s group ¹ looked at a specific set of cancers, those associated with K-Ras mutations (linked to smoking-induced lung cancer). K-Ras activation itself triggers inflammation (for reasons I, at any rate, don’t understand). When K-Ras is activated, as well as inflammation, TRegs move into the area, and presumably reduce the inflammation. Depleting the TRegs (and therefore increasing the inflammation) decreased the number of tumors by 75% — the opposite of what you’d expect if inflammatory RONS were driving tumorigenesis.

A common feature linking smoking induced K-Ras mutations in human lung cancer and preclinical models driven by tobacco carcinogens that cause K-Ras mutations is inflammation. In both cases, the presence of Foxp3+ cells is likely important for limiting the extent of inflammation and tissue damage, albeit at a potential cost of promoting tumorigenesis. ¹

In later-stage tumors the situation became more consistent with other work — getting rid of TRegs reduced the tumors, suggesting that these tumors were depending on TRegs to prevent immune clearance:

Aggressive and invasive K-Ras-induced adenocarcinomas (IO33 and K-RasLA2) remained sensitive to more direct targeting of Foxp3+ cells through a neutralizing anti-CD25 antibody or genetic deletion. This indicates that direct Treg cell depletion strategies that are being evaluated clinically could have therapeutic value in more advanced stages of K-Ras driven lung cancer. ¹

My question here is whether the early inflammation is kind of a red herring. Could the TReg depletion in the early stages be reducing the anti-tumor immune response in a specific way, just as in the later stages of tumor formation? That is, could the TReg depletion lead to a tumor-specific immune response, which prevents tumors from forming? In this case the inflammation could still be driving the tumor formation, but the increase in tumor formation would be outweighed by the simultaneous increase in anti-tumor immunity. I don’t know quite how to test this, but perhaps doing the same experiment in mice lacking, say, CD8 T cells might be interesting. (Such mice should still have the early inflammation and the TRegs, but may have a less effective immune response. It’s not a perfect experiment, though, for reasons that are probably too complex to go into here.)

1. Granville, C., Memmott, R., Balogh, A., Mariotti, J., Kawabata, S., Han, W., LoPiccolo, J., Foley, J., Liewehr, D., Steinberg, S., Fowler, D., Hollander, M., & Dennis, P. (2009). A Central Role for Foxp3+ Regulatory T Cells in K-Ras-Driven Lung Tumorigenesis PLoS ONE, 4 (3) DOI: 10.1371/journal.pone.0005061[↩][↩][↩][↩]

Melanoma antigens

The other day I talked about about resurrecting the antiviral response in HIV patients. ¹ Antiviral T cells in HIV (and other chronic immune responses) become exhausted: After long exposure to antigen, the cytotoxic T lymphocytes (CTL) become dysfunctional, incapable of mounting a potent response to the virus. This exhausted state is correlated with a number of surface flags, especially the molecules PD-1 and CTLA-4. These aren’t merely flags, but rather they actually transmit the signal to become exhausted.  So it turns out that blocking PD-1 reversed the exhaustion, restored  CTL to their youthful vigor, and allowed them to effectively suppress the virus replication. All the monkeys treated with PD-1 blockade survived, whereas most of those left untreated died within a few months.

As I say, exhaustion isn’t unique to HIV. Probably any chronic exposure to antigen tends to cause  T cell inhibition. There’s molecular logic behind this; if you’ve been fighting an infection for many months, you’re probably not winning, and your immune response is probably doing as much damage as the infection would. Or — even worse — you’re not fighting an infection at all, you’re attacking yourself (because of course you can’t eliminate your own antigens). So maybe it’s time to back off a few notches on the attack and try to reach an accommodation with the antigen.

There are a number of cases — probably many cases — where this seems to work well. Rodents that are chronically infected with hantaviruses turn on a regulatory T cell (TReg) type response, shutting own the attack on the virus and letting them become persistent infections. This comes with some cost, but not too much; probably the infected rodents do much better by letting the virus persist, than if they kept trying to fight the infection.

TRegs in skin

There’s another condition when T cells chronically attempt to attack foreign antigen, frequently fail to eliminate it, and become inhibited. This is, of course, cancer. The nature of the CTL inhibition may not be exactly the same as in HIV infections and other CTL exhaustion scenarios, but it’s pretty clear that in general, CTL are not very effective against tumors. After all, most tumors don’t spontaneously regress after a few weeks.

This is probably because when CTL are effective against tumors, that tumor never becomes detectable. In other words, we are only aware of those cancer where CTL are ineffective. (See here (part I) and here (part II) for more detail.) What often happens with tumors, that may be less of an issue with virus infections, is that TRegs become activated and move into the tumor; TRegs shut down aggressive immune responses. As a result, even if you infuse the patient with active anti-tumor cells, or vaccinate and activate the anti-tumor response that way, the anti-tumor response is often quickly shut down by the TRegs and the response never really goes very far.

So can the ineffective T cell response in tumors be reversed, as was done with the ineffective T cell response in SIV? It certainly can — but, as with most anti-tumor immune therapies, it doesn’t work all the time.

With tumors, unlike virus-associated exhaustion, the CTL dysfunction seems to be often associated with the CTLA-4 cell marker. As with PD-1, CTLA-4 isn’t just a marker, it transmits signals into the T cell and actively drives the cells into an inhibited state. (CTLA-4 is probably part of the TReg arsenal, though not the whole of it.) So blocking CTLA-4 in tumor patients has been of intense interest for quite a long time — I think Jim Allison first tried it well over a decade ago². In general the results have been encouraging, but unspectacular. (It seems that immune treatment of cancer is always encouraging but unspectacular. The problem has been to get consistent effectiveness, rather than occasional amazing cures.)

Melanoma blood vessel

This isn’t a safe and innocuous treatment. CTLA-4 is part of the normal immune regulation machinery, and given that, it’s not surprising that CTLA-4 blockade often leads to autoimmunity. In fact, it seems that the more effective the anti-tumor effect is, the more likely the patient is to develop autoimmunity – sometimes quite severe. Compare this to the PD-1 blockade in monkeys, where there wasn’t much autoimmunity, if any.  (Incidentally, before the PD-1 blockade that seemed to work, CTLA-4 blockade has been tried in SIV-infected monkeys.  It didn’t seem to do much.)

A recent paper³ has connected CTLA-4 blockade to the emerging theme of polyfunctionality. As I’ve noted before, it’s become clear over the past couple of years that not all CTL are equal. In HIV infection, polyfunctional CTL — CTL that are capable of producing a wide range of effects, rather than just one or two — are often linked to suppression of the virus. In melanoma patients treated with CTLA-4 blockade, not only were more T cells specific for melanoma antigens present, but those CTL were more likely to be polyfunctional — thus more likely to be effective at destroying the tumor — and those patients were much more likely to have regression of their tumors than in people without CTLA-4 blockade.

So the concept that TRegs — or some other inhibitory effect associated with CTLA-4 — suppress anti-tumor immune responses is likely to be correct, and it seems that at least in some cases it’s possible to override that inhibition and drive T cells to once again attack the tumor effectively. When that happens, cancer can be cured. It’s just a question of being able to do this on a consistent basis. Unfortunately, that’s still the hard part.

1. The actual experiment was done in SIV infected macaques, but of course the hope is that it will translate to the human virus as well.
   [↩]
2. Enhancement of antitumor immunity by CTLA-4 blockade. Leach DR, Krummel MF, Allison JP. Science. 1996 Mar 22;271(5256):1734-6.
   [↩]
3. Yuan, J., Gnjatic, S., Li, H., Powel, S., Gallardo, H., Ritter, E., Ku, G., Jungbluth, A., Segal, N., Rasalan, T., Manukian, G., Xu, Y., Roman, R., Terzulli, S., Heywood, M., Pogoriler, E., Ritter, G., Old, L., Allison, J., & Wolchok, J. (2008). CTLA-4 blockade enhances polyfunctional NY-ESO-1 specific T cell responses in metastatic melanoma patients with clinical benefit Proceedings of the National Academy of Sciences, 105 (51), 20410-20415 DOI: 10.1073/pnas.0810114105
   [↩]

U-118 Glioma cells

In one of the earliest posts I made to Mystery Rays,  I commented on an exciting anti-tumor finding. ¹  Basically, the suggestion was that when chemotherapy of tumors works, it doesn’t actually work by killing all the tumor cells; or at least not directly.  Instead, the authors said, chemo works because the dying tumor cells release adjuvants (immune stimulants),² and what actually eliminates the tumor is the specific immune response to the tumor, driven and enhanced by the immune stimulants.   In other words, the chemotherapy is actually acting like the adjuvant in an anti-cancer vaccine. 

Presumably the reason this form of adjuvant is effective, whereas just giving an adjuvant to a cancer patient is not, is that it’s so tightly linked to the tumor you want the immune system to target.   

In the subsequent year and a half there have been a handful of papers looking more at this phenomenon (not counting the review papers, which have churned out at a great rate). For example, it’s been shown that dying tumor cells are cross-presented efficiently due to HMGB1 and similar immune stimulators.³    Now, there’s a more specific support of the concept. ⁴ 

I don’t know if the Apetoh et al paper was the initial impetus for the latest one, or (more likely) if it was already in progress, but a any rate, unless I’ve missed one, this is the second study that directly looks at HMGB1 effects on tumor regression following tumor cell death, and it reaches basically the same conclusion as the first paper.  This was strictly a mouse study, but they used a highly aggressive brain tumor (glioblastoma multiforme).  As well as chemotherapy they used a gene therapy approach, thus managing to hit four of my Mystery Rays buzzwords (tumor immunity, immunity to viruses, innate adjuvants, and if you’re generous oncolytic viruses as well) all at once.  (What would a Mystery Rays Bingo Card look like?)

Glioblastoma

If the model is correct, then what you really want to do when treating tumors is a double whammy — kill tumor cells, and simultaneously attract in immune cells.  Curtin et al did this by injecting two recombinant adenoviruses directly into the tumor; one rendered the infected cells sensitive to chemotherapy, and the other attracted dendritic cells (DC are important cells for initiating immune responses). Following chemo, the tumors regressed dramatically (about half the mice survived, compared to 100% death without treatment).  And both of the viruses were necessary; without the immune system, killing the cells wasn’t enough, and without tumor cell death, the immune system couldn’t do enough. 

The same treatment in mice without TLR2 did nothing.  TLR2 is an innate immune recognition molecule that triggers inflammation in the presence of (among other things) HMGB1, which is released from dying cells.  This is similar to the Apetoh et al paper, and Curtin et al took the studies a little further by showing that TLR2 had to be on dendritic cells (as opposed to the tumor itself). What’s more, when all the parts were in place (tumor cell death, DC  present, DC expressing TLR2) there was a strong specific anti-tumor cytotoxic T lymphocyte response, while without the TLR2 stimulation there was little such response.

In other words, they’ve firmed up the probable pathway by which chemotherapy eliminated these tumors. The chemotherapy killed cells directly, and the dying cells released HMGB1; the HMGB1 activated dendritic cells in the tumor, by triggering their TLR2 receptor; the activated DC then in turn activated CTL specific for the tumor; and the CTL then completely eliminated the tumor and prevented the remains from regrowing.  (Some of the steps in this pathway remain formally unproven, but the data are consistent with this model.) 

It’s not clear, yet, if this is the universal explanation for successful chemotherapy, or whether chemo sometimes or often works by killing the tumor directly with no requirement for immunity. The authors here looked at several different tumors, expanding on the more correlative data from Apetoh et al., and so far all the cancers that have been checked fall into the immune clearance category.  So, while it’s a long, long way from the bedside, it’s certainly encouraging. 

In conclusion, the results reported provide compelling evidence for the role played by HMGB1 in mediating the efficacy of antiglioma therapeutic regimes that are based on tumor cell killing strategies … [C]ancer immunotherapies coupled with effective cell killing modalities may be necessary to achieve therapeutically relevant antitumor efficacy. ⁴

1. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Apetoh et al. Nature Medicine 13, 1050 – 1059 (2007)  doi:10.1038/nm1622

   Also this review:

   Apetoh, L., F. Ghiringhelli, A. Tesniere, A. Criollo, C. Ortiz, R. Lidereau, C. Mariette, N. Chaput, J. P. Mira, S. Delaloge, F. Andre, T. Tursz, G. Kroemer, and L. Zitvogel. 2007. The interaction between HMGB1 and TLR4 dictates the outcome of anticancer chemotherapy and radiotherapy. Immunol. Rev. 220:47-59.

   and this one:

   Apetoh, L., A. Tesniere, F. Ghiringhelli, G. Kroemer, and L. Zitvogel. 2008. Molecular interactions between dying tumor cells and the innate immune system determine the efficacy of conventional anticancer therapies. Cancer Res. 68:4026-4030.

   and this review from a different group:

   Campana, L., L. Bosurgi, and P. Rovere-Querini. 2008. HMGB1: a two-headed signal regulating tumor progression and immunity. Curr. Opin. Immunol. 20:518-523.[↩]

2. Specifically, the dying cells release a protein called HMGB1, which has been shown to be an endogenous adjuvant[↩]
3. Brusa, D., S. Garetto, G. Chiorino, M. Scatolini, E. Migliore, G. Camussi, and L. Matera. 2008. Post-apoptotic tumors are more palatable to dendritic cells and enhance their antigen cross-presentation activity. Vaccine 26:6422-6432.[↩]
4. James F. Curtin, Naiyou Liu, Marianela Candolfi, Weidong Xiong, Hikmat Assi, Kader Yagiz, Matthew R. Edwards, Kathrin S. Michelsen, Kurt M. Kroeger, Chunyan Liu, A. K. M. Ghulam Muhammad, Mary C. Clark, Moshe Arditi, Begonya Comin-Anduix, Antoni Ribas, Pedro R. Lowenstein, Maria G. Castro (2009). HMGB1 Mediates Endogenous TLR2 Activation and Brain Tumor Regression PLoS Medicine, 6 (1) DOI: 10.1371/journal.pmed.1000010[↩][↩]

“Jennerex” virus (green) replicating within a tumor mass.

Viruses infect cells, and quite often (depending on the type of virus) destroy the cell they’re infecting. Usually having your cells destroyed is a bad thing for the host, because you need those cells. But there are some cells that you don’t want to have, and it might be convenient to have a controlled virus destroy those cells for you.

The obvious example would be cancer cells, and in fact there’s a flourishing research industry that is trying to harness the destructive power of viruses to eliminate tumors. The trick, of course, is to have the virus infect only the cancer cells, not the normal healthy cells you want to keep; and the general approach is to take advantage of some of the common features of viruses and cancers. Cancers have to mutate to overcome some of the same cellular controls that viruses do (both viruses and cancers need to overcome the regulation of uncontrolled genome replication, for example). If you cripple some of these viruses, then, they can’t replicate in normal cells, but can replicate in, and destroy, cancer cells that have mutated the appropriate pathway.

(I talk about oncolytic viruses in more detail here.)

As often happens with these intriguing cancer treatments, oncolytic viruses seem to work sometimes, and not to work sometimes, and it’s not always clear why not. In a paper in PLoS One the other day,¹ Wodarz and Komarova attempt to come up with a way of predicting which tumors will and will not respond to oncolytic virus therapy, using a computational approach.

As I think I’ve said before, I’m excited by the concept of computational biology. (I’m not so much talking about bioinformatics as such here, but rather about attempts to model AND PREDICT complex biological processes.) But I’ve been kind of disappointed by some of the process. It’s seemed to me that when simple processes are modeled we haven’t really learned very much new, and when complex processes are modeled the assumptions are often too simplistic to make a reasonable prediction. I don’t think we’re at a point where we can usefully model an immune system, for example.

Growth of cancer in a mouse in the presence of  oncolytic virus 1

However, I do think there are a class of problems in the middle where the approach is more successful, and though I’m not really able to critically assess their results I think over the years Dominik Wodarz has done a good job of identifying these problems. Questions like the emergence of drug resistance in cancer,² effects on vaccination on HIV,³ and so on seem like the kind of problem where mathematical analysis can actually get a handle on the issues and help guide research to some extent. Again, I don’t feel that I can really judge the results, but I like the approach. What’s more, Wodarz seems to at least consider experimental evidence in his analyses, which isn’t always the case in these computational things.

(I don’t think it’s a coincidence that many of the questions he’s asked are microcosmic versions of ecological issues.  My impression is that population biology has a much longer and more successful history of mathematical analysis than do cell and molecular biology.)

This particular paper leads to a conclusion that, once reached, seems fairly obvious in hindsight, but it’s one I haven’t seen explicitly made before. (I am not in the field, and it may be taken for granted.  That said, one hallmark of a successful prediction is that everyone immediately says they knew it all along.) Briefly, tumor growth rate per se turn out to be relatively unimportant, and growth patterns are important. If the cancerous cells are relatively spread out in the tumor, then an oncolytic virus has a good chance of eliminating the tumor; whereas if the cancer cells are in clumps, the virus is much less effective. This is simply because in the clumpy masses most of the infected cells are contacting already-infected cells, and the only route to reach new targets is from the surface of the clump, so spread is inefficient.

In one group, virus growth is relatively fast because the infected cells are dispersed among the uninfected cells rather than being clustered together. In this case most infected cells contribute to virus spread. In these models, there is a clear viral replication rate threshold beyond which the number of cancer cells drops to levels of the order of one or less, corresponding to extinction in practical terms. … In the other category, infected cells are assumed to be clustered together to some degree in a mass, which might be realistic for solid tumors. In this case, only the infected cells located at the surface of the cluster contribute to virus spread because they are in the vicinity of uninfected cells. … In this scenario, virus therapy is more difficult. ¹

1. Dominik Wodarz, Natalia Komarova (2009). Towards Predictive Computational Models of Oncolytic Virus Therapy: Basis for Experimental Validation and Model Selection PLoS ONE, 4 (1) DOI: 10.1371/journal.pone.0004271[↩][↩][↩]
2. Drug resistance in cancer: principles of emergence and prevention. Komarova NL, Wodarz D. Proc Natl Acad Sci U S A. 2005 Jul 5;102(27):9714-9.[↩]
3. Immunity and protection by live attenuated HIV/SIV vaccines. Wodarz D. Virology. 2008 Sep 1;378(2):299-305.[↩]