Monday, April 9, 2012

CD47 Antibody: A Promising Breakthrough

Introduction & Background
Irving Weissman (8)
On the first day of class, Dr. Islas mentioned that our best defense against cancer is not through medical interventions (i.e. burning, radiation, or resection), but rather our own immune system.  The reason we so susceptible to this disease, however, is the fact that cancer has evolved mechanisms to avoid our immune system.  Over the weekend, I came across a study conducted by a group of Stanford scientists who have discovered a way to disable one of these “mechanisms,” so that our body’s inherent defenses can better detect and eliminate cancer cells.
The senior author of the study is Irving Weissman, MD, a professor of pathology at the Stanford University School of Medicine.  The research was published online on March 26, 2012 in the Proceedings of the National Academy of Sciences.
Weissman and his team discovered that human tumors transplanted into laboratory mice disappeared or shrank when they treated the animals with a single antibody (1).  The antibody works by masking a protein flag on cancer cells, called CD47, which protects the cancer cells from macrophages and other cells involved in immune defense.  Weissman achieved the findings for a wide range of human cancers including, breast, ovarian, colon, bladder, brain, liver and prostate.  Their discovery is the first antibody treatment shown to be effective against human solid tumors (2).  With such dramatic results, the investigators hope to begin phase-1 and phase-2 human clinical trials within the next few years.  Before delving into exactly how the scientists went about their study, let’s take a step back and talk about CD47.
CD47 is a transmembrane protein that cells use as a mechanism to protect themselves from phagocytosis (1).  Normally, CD47 is expressed on the surfaces of circulating blood stem cells to protect them from engulfment by macrophages (3).  Macrophages are part of the body’s innate (non-specific) immune response.  They patrol the body looking for foreign or diseased cells, but they sometimes make mistakes and bind to the wrong targets (2).  When a cell with CD47 binds to signal regulatory protein-a (SIRP-a), a protein expressed on macrophages, it initiates a signaling cascade that inhibits phagocytosis.  Therefore, the binding of CD47 to SIRP-a prompts macrophages to release cells they may have grabbed accidentally before the cell is phagocytosed (1).
Previous work in Weissman’s lab determined that human leukemia and lymphoma cells tricked the system by expressing CD47.  By mimicking normal blood stem cells, the cancer cells were able to inhibit the macrophage’s killing instinct and escape unharmed.  At the time, Weismann nicknamed CD47 as the “don’t-eat-me signal.” (3)
Weissman and his team took this initial CD47 finding and went on to discover that blocking CD47 with an antibody can cure some cases of human non-Hodgkin’s lymphoma in mice (4).  But, it is his most recent findings in the current study that has huge implications for a potential cancer therapy against human solid tumors.

CD47 is Expressed on Solid Tumor Cells
Weissman and his team collected samples from a variety of human tumors including, breast, ovarian, colon, bladder, brain, liver and prostate (2).  Tissue specimens were obtained from consented patients at Stanford Hospital (1).  Using flow cytometry researchers determined that nearly every human cancer cell they examined expressed CD47 (Fig 1a).  CD47 was also detected on some normal (noncancerous) cells (Fig 1a).  Using quantitative flow cytometry, Weissman also found that tumor cells, on average, expressed 3.3-fold more CD47 than corresponding normal (noncancerous) cells (Fig 1b).

From this, the Stanford researchers showed that CD47 is not only present on the surfaces of blood cancer cells (leukemia, lymphoma), but also expressed on the surfaces of a wide variety of human solid tumors.

Anti-CD47 Antibodies Inhibit Tumor Growth

Weissman and his team then implanted different human cancer cells into similar locations in the bodies of immunodeficient NSG (NOD scid gamma) mice.  NSG mice lack B, T, and natural killer cells, but retain macrophages (1).  Prior to injection of the human cancer cells, the cells were transduced with a GFP and luciferase-encoding lentivirus.  This allowed researchers to use bioluminescence to measure tumor growth.  Once the tumors were well established (after two weeks or longer), they treated the mice with the blocking anti-CD47 antibody (2).  Control mice were injected with IgG or a non-blocking anti-CD47 antibody (1).


Ovarian Cancer
            Ovarian cancer cells were placed in the abdomen of mice.  Treatment with anti-CD47 antibodies inhibited tumor growth as seen by bioluminescence imaging (Fig 2b).  Additionally, mice given the anti-CD47 antibody exhibited a significant increase in survival (Fig 2c).

Breast Cancer
            Breast cancer cells were placed in the mammary fat pads of mice.  After 8 weeks, IgG-treated (control) mice developed large tumors in contrast to anti-CD47 antibody treated mice that showed no palpable tumors (Fig 3a).  After week 8, researchers stopped infusing the mice with the anti-CD47 antibody treatment.  The mice were monitored for an additional four months with no signs of recurrence.  This suggests that the anti-CD47 antibody therapy completely eradicated all breast cancer cells, including cancer stem cells.  In a second breast cancer sample that was tested, anti-CD47 antibody treatment again inhibited tumor mass significantly (Fig 3b).

Colon Cancer
            Colon cancer cells were placed in the back of mice.  Treatment with anti-CD47 antibodies inhibited tumor mass significantly (Fig 4).

Glioblastoma (Brain)
Glioblastoma cells were placed in the left hemisphere of mouse brains.  After 8 weeks, anti-CD47 antibody treated mice showed a significant reduction in tumor growth (Fig 5a).  Additionally, the anti-CD47 antibody may have completely eliminated tumors in some mice as seen by bioluminescence imaging (Fig 5b).

From these results, Weissman concluded that “anti-CD47 antibodies can dramatically inhibit the growth of human solid tumors by blocking the ability of CD47 to transmit the ‘don’t-eat-me’ signal to macrophages.” (1)

Anti-CD47 Antibodies Prevent Tumor Metastasis
            Next, Weissman and his team wanted to know if the anti-CD47 antibody would be effective against a tumor that was highly aggressive and very likely to metastasize.

Bladder Cancer
            To examine this question, Weissman and his partners used human bladder cancer samples that easily metastasized to the lymph nodes and lungs after implantation into the back of NSG mice.  Once the tumors were well established, they treated the animals with the anit-CD47 antibody (1).  Using gross examination, they found significantly less metastases in the lymph nodes and lungs of mice treated with the anti-CD47 antibodies compared to IgG-treated (control) mice (Fig 6a, b, c, d).
            The results indicate that anti-CD47 antibodies may be effective at inhibiting the ability of tumors to metastasize.

Ultimately, Weissman and his team showed that the CD47 protein is a legitimate and promising target for human cancer therapy.  To relate this blog back to what Dr. Islas told us on the first day, Weissman’s discovery is one that eliminates a specific mechanism used by cancer cells to evade our immune system, thereby allowing our own body’s natural defenses to do their job!

Questions & Thoughts
After reading the paper, a few questions and thoughts came to mind:
1)       When using flow cytometry to measure expression of CD47 Weissman found that CD47 was indeed expressed on some normal (noncancerous) cells.  If this is the case, how are other normal cells in the body prevented from macrophage phagocytosis when exposed to the anti-CD47 antibody treatment?
Possible Answer: In addition to the CD47 “don’t-eat-me” signal, many solid tumors also display a calreticulin (CRT) “eat me” signal.  This means that when anti-CD47 antibodies bind to CD47 on cancer cells, only the CRT signal is exposed.  Display of the CRT “eat me” protein on cancer cells leads to their subsequent phagocytosis by macrophages.  Normal cell populations, however, do not have CRT and are therefore not depleted when exposed to a blocking anti-CD47 antibody (5).  The fact that normal cells are not affected by the anti-CD47 antibody is just one factor that makes Weissman’s finding so promising.  Unlike the toxic effects of chemotherapy on our normal, healthy cells, patients who undergo the anti-CD47 antibody therapy may hopefully exhibit fewer side-effects.

2)       The previous “answer” to question #1, leads to the following question: Why would cancer cells even bother to express the CRT “eat me” signal if it can potentially cause their destruction?
Possible Answer: Perhaps the CRT “eat me” signal confers some benefit to cancer cells.
Creating my own Experiments: Although I am not sure if this question has been explored, I can think of two ways that could help to define the function of the CRT protein on cancer cells.  The first, involves knocking-out the gene that encodes the CRT protein and observing the changes that occur, if any, in CRT-mutant cancer cells compared to CRT-positive cancer cells.  The second experiment I envision, entails creating an anti-CRT antibody, similar to the anti-CD47 antibody Weissman used, to block the CRT "eat me" protein on the outside of cancer cells.  While we would expect that blocking the CRT tag would provide cancer cells the luxury of going undetected by macrophages of the immune system, this would also allow us to monitor any additional effects of inhibiting CRT on cancer cells.  Of course, given what we know from question #1 (above) along with Weissman's findings, we would not want "real" cancer cells to stop expressing the CRT "eat me" protein and thus rendering the anti-CD47 antibody ineffective.  At this point, from our perspective as scientists, it seems that expressing the CRT protein is a disadvantage to cancer cells rather than an advantage.

3)    Why did Weissman use immunocompromised NSG mice?
Answer: My first thought in answering this question was that perhaps Weissman used NSG mice because cancer patients are “sick,” in that often they lack a fully competent immune system.  By using NSG mice, Weissman and his team were attempting to mimic the level of immunity that would be expected of human patients with tumors.  BUT, after reading chapter 3 in our textbook, The Biology of Cancer, I found a better/”the right” answer!  Immunocompromised mice, such as the NSG mice Weissman used, lack fully functional immune systems and are therefore relatively receptive of engrafted cells from genetically unrelated sources, including cells from a foreign species.  In this study, researchers needed the NSG mice to ensure that the human tumor sample spawned in the mice recipient.  If Weissman had used “healthy” mice instead, then the fully functional immune system of those mice would most likely have eliminated the human tumor transplant (pg. 70-71 of 6).

4)    After reading the Hallmarks of Cancer: The Next Generation there seems to be conflicting information regarding the immune response and its role in tumorigenesis.  In their article, Hanahan and Weinberg state that tumors are often densely infiltrated by cells of both the innate (non-specific) and adaptive (specific) branches of the immune systems.  Contrary to what we would expect, the authors said there is substantial evidence to indicate that this infiltration or immune response aids the growth of cancer rather than stopping its progression.  The data suggests that "immune cells – largely of the innate immune system – have tumor-promoting effects on neoplastic progression." (7)  Although the literature is certainly conflicting, I think Weissman and his team were successful in demonstrating that with assistance (i.e. anti-CD47 antibody injection) the immune system can do what we would hope, which is to help eliminate cancer and prevent tumorigenesis.

5)       As young biologists with potentials for careers in research, I think it’s important to acknowledge the progression of Weissman’s research.  In July of 2009, he discovered the CD47 molecule on the outside of leukemia stem cells (3) and now, less than three years later, he has developed and tested an antibody against the CD47 protein that has the potential to become a therapeutic treatment against human solid tumors (1).  Weissman’s incredible progress is a primary example of science at its best.  It is the classic but rare tale, of how one scientific finding can lead to next, which can ultimately lead to huge, impactful discoveries.

6)       This may be useful information for groups interested in doing their cancer project on 1) the role our immune systems play in cancer growth or 2) new, cutting-edge cancer therapies.

Thanks for reading!