This web page was produced by Zack Perfect as an assignment for an undergraduate course at Davidson College


Integrin a 3b 1(CD 49c/29) Is a Cellular Receptor for Kaposi’s Sarcoma-Associated Herpesvirus (KSHV/HHV-8) Entry into the Target Cells

Shaw M. Akula, Naranatt Pramod, FuZhang Wang, and Bala Chandran

Department of Microbiology
Molecular Genetics and Immunology
The University of Kansas Medical Center
Kansas City, Kansas 66160

Cell (2002) Vol. 108, 407-419.


             This paper investigates the mechanism by which the Human herpes virus-8 (HHV-8) is able to attach to a target cell and ultimately infect it. HHV-8 is associated with the disease Kaposi’s Sarcoma, most commonly found in HIV patients in which AIDS has arisen. Through a series of experiments the authors determine that a type of cell receptor known as an integrin is the primary receptor the virus binds to when infecting. Specifically, they determine it is a a3b1 receptor through multiple techniques which include antibodies, immuno-precipitation, Western blots and FACS, among others. The determination of the cell receptor would provide valuable knowledge that could benefit Kaposi’s Sarcoma research and possibly lead to improved treatments (i.e. this is important, so let’s get on with it and see what they found).

Figure 1: This figure merely demonstrates their ability to determine whether HHV-8 infection has taken place and then quantify it through cell counts. 1A and D used mock cells that were not infected, providing the control. Immuno-florescence was used in 1A-C by attaching a Green Florescent Protein onto the HHV-8 virus, infecting, then staining. In 1D-F infection was detected through Mab antibody to reveal the presence of the ORF73 protein. As expected, the two control samples showed no color or stains, while the other four do. These procedures will appear later in the paper and the ability to properly perform them was demonstrated in this figure.

Figure 2: This figure is out to suggest the RGD-dependent nature of the HHV-8. Figure 2A uses four synthetic RGD blockers to determine which one is the most effective and reaffirm that a structure similar to RGD can reduce infection. Overall GRGDTP was the most effective, inhibiting about 80% of the infectivity, followed by RGD and RGDgB-n1, while an unrelated GRGESP served as a control and showed the expected minimal effect, increasing infectivity by only 10%. Figure 2B demonstrated the same trend as 2A but using antibodies. They used rabbit antibodies that bound to the RGDgB-N1 region, preventing the HHV-8 from binding and causing a 50% increase in inhibition. A positive control used was the rabbit anti-bG IgB antibodies, known (for an unexplained reason) to allow infectivity. Two other antibodies were synthetically manipulated so that they were missing the RGD motif, and as expected they were not effective at inhibition. This demonstrates the presence of a surface protein on the virus that can be bound by an antibody. Figure 2C tested whether the receptor on the body cell was integrin in nature, which they hypothesized. To test this they exposed HHV-3 infected cells to different drugs, with only Fibronectin known to affect integrins. The results were convincing, the Fibronectin inhibiting nearly 90% of HHV-8 activity at the highest concentration, while second most successful was 10%. This supports that the receptor is integrin in nature.

Figure 3: There are many types of subunits associated with integrin. The researchers hypothesized that it was a 3 and b 1 that were the primary components, and specifically were paired together to form the true receptor structure. In 31 they used anti-integrin antibodies that effectively inhibited infectivity. The predominant three were a 3, b 1, and a 2b 1. The first two are what they expected to find and they explained the a 2b 1 as being related purely to the presence of the b 1’s interaction, which doesn’t make sense because a 5b 1 was also tested and didn't show similar trends. To reiterate this data they repeated the experiment in Figure B with the antibodies to the three molecules showing high inhibitions, while the control/unrelated antibody, showed little affect. The most powerful inhibitor was a combination antibody that combined all three, making it more likely the receptor is blocked. Finally in 3C, they compared the inhibition of the a 3b 1, the proposed receptor, to the unlikely, but similar combination receptor, a 5b 1. The difference was quite stark, with the a 3b 1 inhibiting 90%, while the a 5b 1 being nearly ineffective. These figures support that a 3 and b 1, especially when grouped, are physical portions of the integrin receptor on the body cell.

Figure 4: A-D show the number of HFF (adult human dermal microvascular endothelial) cells measured by FACS. A is the anti-mouse IgG, which establishes the normal number of cells. The numbers responding to the antibody get progressively larger as the a Vb 1, a 3, and b 1 are used, showing the presence of these receptors in the HFF cells, which are commonly attacked by HVV-8. Figure 4E is similar in that it tests levels of integrins, but it uses a variety of target cells and more integrins. The data are similar, with a 3 b 1, and a 5b 1 appearing in nearly 90% of the cells while the others were found at significantly lower levels.

Figure 5. 5A is a flow cytometric analysis of a 3 integrin using CHO-B2 (hamster) cells. 5A1 is the control, using cells transfected with an unrelated pCNDA3 plasmid. The other two CHO-B2 cells were transfected with two lines B3 and D5 (the B3 expressing four times more of the human/hamster a 3b 1 clone). As predicted, the control number of cells infected was 4% while the D5 was 23% and B3 was 95%. Figure 5B measures infection of CHO-B2 cells by the D5 and B3, visualized by GFP and ORF 73 protein expression. As expected, more cells are seen in B3. Figure 5C shows the infectivity of different antibodies, anti-a 3 or anti-b 1, on the CHO-B2 cells. Once again the CHO-B2 with just an unrelated pCDNA serves as the control, and all three bars are low. The D5 and B3 are increasingly more infective, as would be expected. In both, the lack of either antibody is high, which serves as a control because nothing binding the receptors on the infected cells should produce little prevention of infectivity. The anti-a 3 antibody is much lower, indicating that there is binding to the receptor preventing the virus, and the anti-b 4 is nearly the same infective as neither being there, indicating that while an integrin is there, it is not important. This experiment is similar to the ones that determined which were the important integrins, but this demonstrates in hamster cells in which the disease receptor was experimentally introduced.

Figure 6: 6A shows that it is the gB portion of the HHV-8 that specifically binds with a 3 and b 1 receptors. 14 lanes are exposed to a variety of conditions including with or without anti-gB antibody and varying anti-integrins. When run on a Western blot the only four bands seen were in the a 3 and b 1, but only with the gB, indicating its importance. Then in 6B they tried to determine whether the HHV-8 interacted with the surface receptors during or post attachment. They measured different reagents ability to inhibit thymidine labeled HHV from binding to the cells. For reasons that were not explained, Heparin must only indicate infection during attachment because this is the conclusion they draw and Heparin is significantly higher than any of the other reagents in binding inhibition. The mechanisms behind this graph were somewhat ambiguous.

Figure 7: This figure investigates the proposed association between HHV-8 and signaling molecules commonly released as the result of integrin binding. Two of the choice signaling proteins were FAK and vinculin. 7A shows evenly distributed FAK and viculin throughout a mock cell, serving as a negative control. In HHV-8 infected cells and LPA showed spotty, highly concentrated sections of the two proteins, indicating FAK and vinculin were being utilized. LPA must mimic successful initiation of chemical pathway and thus serve as a positive control (again, not explained, but possibly common knowledge). The HHV-8 cells show similar trends, indicating that indeed the virus binding to the integrin and activating FAK and vinculin is mostly likely the mechanism of infection. 7B, C and D were Western blots in which the presence of FAK was measured and actin’s presence was used a control. 7B used varying levels Phospho-FAK at increasing time intervals to indicate that indeed the FAK is being phosphorolated. Similarly in 7C, a partial HHV-8 inhibitor (obviously not effective enough or they probably wouldn't be doing this study) in decreasing amount led to less Phospho-FAK in progressive lanes. In 7D anit-gB antibodies were used to prevent infection and the same lack of Phopho-FAK was seen. These figures demonstrate that successful HHV-8 infection is directly related to the amount of FAK in the body cell.

            Both the sequence of the research and its presentation in the paper was very well executed. Although the researchers may have suspected various types of integrins from the beginning, they went ahead and proved there was a cell receptor and that it seemed to be integrin in nature before they investigated the specifics and proceeded to confirm the actual mechanisms and signaling pathway induced by attachment of the virus. A final step that could have been taken and provided the most convincing evidence would be a knockout mouse with target cells lacking the a 3b 1 integrin. If this is possible, with creating an otherwise normal mouse, then in theory the knockout could be injected with the HHV-8 and if it exhibited no signs of Kaposi’s Sarcoma, the proverbial "nail in the coffin" will have been achieved. There could exist, and the authors mention this at the end of the paper, other receptors that work with this primary a 3b 1 receptor. It would be beneficial to search for these receptors in the future and determine whether they alone can cause infection or merely work to enhance primary receptors. Another area of future investigation could be possible mechanisms, and ultimately drugs, that target inactivating or eliminating the functional binding site of HHV-8. In particular, the study could involve the RGD motif that this paper has shown to be crucial to virus binding and delivery. Overall, this appears to be an important finding that could lead to improved treatments of Kaposi’s Sarcoma, particularly in the HIV infected.