A new role for CD28 in the survival of autoreactive T cells in the periphery after chronic exposure to autoantigen

A new role for CD28 in the survival of autoreactive T cells in the periphery after chronic exposure to autoantigen
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A new role for CD28 in the survival of autoreactiveT cells in the periphery after chronic exposure toautoantigen Jian-Xin Gao, Xing Chang, Xincheng Zheng, Jing Wen, Lijie Yin, Peishuang Du,Pan Zheng  and  Yang Liu Division of Cancer Immunology, Department of Pathology, Ohio State University Medical Center,Columbus, OH 43210, USA Keywords  : cell-surface molecules, repertoire development, T lymphocyte AbstractRecent work demonstrates that costimulatory molecules play a critical role for clonal deletion ofautoreactive T cells in the thymus. The role of CD28 in the survival of autoreactive T cells in theperiphery, however, has not been reported. Here we demonstrate that while mutation of the  CD28  gene consistently increased the burden of autoreactive T cells in the thymus, such an increase wasnot always found in the periphery, as the CD28(   /   ) autoreactive T cells disappeared in the spleenover a period between 4 and 10 weeks. The disappearance of autoreactive T cells associates witha diminished induction of Bcl-2 protein by the self antigen and an increased proportion of apoptoticcells in the periphery. Moreover, the elimination of autoreactive T cells in the periphery requireschronic stimulation by the self antigen, as adoptive transfer analysis revealed no enhancement ofapoptosis in CD28(   /   ) T cells in antigen-bearing hosts over a 3 day period. Thus, CD28 playsa significant role in both clonal deletion and survival of autoreactive T cells after chronic exposure toautoantigens, resulting in opposite effects on the burden of autoreactive T cells.Introduction The repertoire of autoreactive T cells is controlled by theirclonal deletion and their survival in the periphery. Several linesof previous studies suggest that the costimulatory pathway isinvolved in T cell clonal deletion. First, in thymocyte sus-pension cultures, a role for B7-1/2 and CD28 has been welldocumented. Thus, anti-B7-1/2 antibodies and fusion proteinswith similar activity block the death of antigen-inducedthymocytes  in vitro   (1–3). Agonistic anti-CD28 mAb was alsofound to promote antigen-induced deletion of T cells (2,4,5).Correspondingly, transfection of B7-1 into antigen-bearingcells promotes antigen-induced death of immature thymo-cytes (6). Second, Page reported that anti-B7-1 and anti-B7-2prevented clonal deletion of MHC class II-restricted trans-genic T cells (AND), induced by either cognate antigenicligand or allogeneic MHC (H-2 s ) (7), in the thymic organculture. Third, several  in vivo   analyses suggest that in additionto TCR engagement, costimulatory molecules may contributeto the efficacy of clonal deletion. Thus, Sprent and colleaguesshowed that deletion caused by  in vivo   injection of low dosesof antigen is affected by CD28 (8). Li and Page have reportedthat among the F2 mice of the BALB/c 3 CD28(   /   )B6 mice,the CD28(   /   ) mice have significant, although small, defectsin the deletion of V b 5 + or V b 11 + CD4 Tcells (9). Interestingly,CD28(   /   )micehavealmosttwicethenumberofthymocytes,and the CD28(   /   ) thymocytes appear more resistant to anti-CD3 induced cell death  in vivo   (10). Should the increase ofthymocyte numbers be attributed to the decreased clonaldeletion, one would expect a more substantial effect of CD28deficiency on the accumulation of self-reactive Tcells.Wehaverecentlyrevisitedtheissueofwhethercostimulatorymolecules B7-1 and B7-2 are involved in clonal deletion byperinatalblockadeofB7-1 andB7-2 (11). We reportedthat thistreatment can block clonal deletion of Tcells specific for endo-genousVSAgs.Inaddition,inmicetransgenicforboththetumorantigenP1AandP1A-specificTcells,wefoundthatanti-B7-1/2mAbs resulted in drastic reductions of T cell clonal deletion.Morerecently,similarresultswerereportedinmicewithtargetedmutations of B7-1/2, CD28 and CTLA-4 (12). Together, theseresults established a more general function of CD28 in clonaldeletion of autoreactive cells than what was perceived. Correspondence to  : Y. Liu; E-mail: Transmitting editor  : C. Terhorst  Received   4 May 2004,  accepted   9 July 2004 International Immunology, Vol. 16, No. 10, pp. 1403–1409   ª 2004 The Japanese Society for Immunology  doi:10.1093/intimm/dxh141 Advance Access publication 16 August 2004   b  y g u e  s  t   onM a  y2  ,2  0 1  6 h  t   t   p :  /   /  i  n t  i  mm . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om  A role for CD28 in the survival of autoreactive Tcells has notbeen reported. Here we report a new role for CD28 in thesurvival of autoreactive T cells in the periphery after chronicexposure to self antigen. Our results demonstrate that thesurvival of autoreactive Tcells in the periphery is a checkpointregulated by the costimulatory pathway. Methods Experimental animals  CD28-deficient and wild-type BALB/c mice were purchasedfrom the Jackson Laboratories (Bar Harbor, ME). TransgenicBALB/c mice expressing tumor antigen P1A under the Id2promoter and E l  enhancer and/or TCR specific for tumorantigen P1A (P1CTL) have been described elsewhere (13).The CD28(   /   )P1CTL + mice have also been reported (14).These mice were all backcrossed to BALB/c backgroundfor more than 10 generations and were used to produceCD28(   /   )P1CTL + P1A + mice. All mice weremaintained in theUniversity Laboratory Animal Research Facility at the OhioState University under specific-pathogen-free conditions. Flow cytometry  Single-cell suspensions of thymocytes and splenocytes wereprepared and stained for expression of CD4 and CD8 incombination with cell maturation marker HSA (heat-stableantigen), P1CTL TCR marker V a 8, and viral superantigen(VSAg)-reactive Tcell markers V b 3, 5, 8, 11 and 12. Anti-Bcl-2antibody was used for intracellular staining of Bcl-2 expres-sion, while annexin V was used to determine cells undergoingprogrammed cell death. The annexin Vand mAbs conjugatedwith various fluorescent dyes were purchased from BD-PharMingen(SanDiego,CA).Insomeexperiments,thestainedcells were analyzed by three or four-color flow cytometry. Adoptive transfer of P1CTL T cells into P1A transgenic mice  Purification of CD8 T cells from P1CTL transgenic mice andtheir labeling with CFSE have been described (15). The CFSE-labeled WT and CD28(   /   ) P1CTL cells were injectedintravenously into syngeneic mice that express P1A asa transgene. At 20, 42 and 66 h after adoptive transfer, thespleen cells were harvested and analyzed for the number ofdivisions based on their CFSE intensity and for otherparameters by flow cytometry. Results CD28 is required for both clonal deletion and survival of antigen-specific T cells  WehavepreviouslyproducedmicethatexpresstumorantigenP1A in the thymus (13). The over-expressed antigen causesclonal deletion of the T cells expressing the transgenic TCR(P1CTL) from a P1A-reactive CTL clone. The deletion can beinhibited by treatment with anti-B7-1 and B7-2 mAbs (11). Weproduced CD28(   /   ) and CD28(+/+) mice that were trans-genic for both P1A and P1CTL, and compared the numbersand subset distribution of the transgenic T cells. Among theCD28(+/+) mice, transgenic expression of the tumor antigenP1A resulted in about a 2-fold reduction in the number ofthymocyte(Fig.1),aswehavereported(11).Targetedmutationof CD28 significantly increased the number of thymocytesin mice transgenic for both P1A antigen and its specific TCR(Fig. 1a).Inthe CD28(   /   )background,the numberofthymo-cytes was also lower when the tumor antigen was expressed,although this reduction is not statistically significant ( P   = 0.07).Wehave reported that clonal deletion of P1A-reactiveTcellsbytheP1AantigenresultedinspecificreductionofCD4 + CD8 + T cells (11). As shown in Fig. 1(b and c), on average, CD28(+/+)P1ATg + P1CTL + mice have ~40% of CD4 + CD8 + T cells,which is sharply reduced in comparison to ~70% found inCD28(+/+)P1ATg  P1CTL + thymi. The difference betweenP1ATg + and P1ATg  mice very much disappeared in theCD28(   /   ) mice. CD28(+/   ) mice had an intermediate levelof clonal deletion and significant clonal deletion was observedin most but not all mice (Table 1). The function of CD28 incausing a reduction in the number of total thymocytes and inreduction of CD4 + CD8 + transgenic T cells is consistent withthe idea that CD28 may play a role in negative selection ofTcells by the P1A antigen over-expressed in the thymus.Interestingly, CD28-deficiency also resulted in a significantincrease in the total number of thymocytes among mice trans-genicforP1CTLbutnotP1A.SincetheP1Aisaselfantigenwithexpression in testis and lymphoid organs, such as spleen (13)and thymus (16), it is likely that ablation of CD28 inhibitedthe negative selection mediated by the endogenous P1A,reportedly expressed in the peripheral antigen-expressingcells in the thymus (16).Paradoxically, the rescue of T cells in the thymus fromnegative selection did not always result in an increasednumber of T cells in the periphery. A clear-cut example ispresented in Fig. 2(a). Two 7-week-old littermates that differedinP1ATghadsimilarthymocytesubsetdistributionswhenCD4and CD8 markers were used. Furthermore, among theCD4  CD8 + T cells, the proportion of HSA low cells was alsocomparable (Fig. 2b). Since expression of HSA is a marker formaturity of the single-positive thymocytes (17), the amount ofmature T cells in the two thymi must be comparable. Theseresults further underscore the notion that the deletion imposedby transgenic expression of P1A can sometimes be com-pletely ablated by targeted mutation of the CD28. The onlydifference that we observed was the increased proportion oflarger cells, perhaps reflecting the stimulation these cellsreceived from the transgenic P1A antigen. Surprisingly,despite normal development of T cells in the thymus,CD28(   /   ) mice had very few CD8 T cells in the spleen.Figure 2(c) depicts the proportion of T cells in the spleenof CD28(   /   ) mice that were either P1A + P1CTL + orP1A  P1CTL + . Although the two mice have an almost identicalproportion of CD4  CD8 + thymocytes (Fig. 2a and b), therewas an almost 7-fold difference in the proportion of matureV a 8 + CD8 Tcells in the spleens (Fig. 2c). The CD8 Tcells werelarger in the spleen of P1ATG + mice.A comparison of 4- and 7-week-old CD28(   /   ) micerevealed that, even though the numbers of mature transgenicTcells were comparable among mice of the two different agegroups, the number of transgenic T cells in the peripheral isreduced by  > 5-fold in 7-week-old mice (Table 2). Theseresults support the notion that disappearance of autoreactivecells may require chronic exposure to autoantigens. 1404  CD28 and the burden of autoreactive Tcells    b  y g u e  s  t   onM a  y2  ,2  0 1  6 h  t   t   p :  /   /  i  n t  i  mm . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om  CD28 promotes clonal deletion and the survival of VSAg-reactive T cells  BALB/c mice have insertion of mouse mammary tumor virus 6,8 and 9 and normally delete Tcells expressing V b 3, 5, 11 and12 (18). We have shown that in BALB/c mice, perinatalblockade of anti-B7-1 and anti-B7-2 significantly inhibitedclonal deletion of VSAg-reactive T cells. To study thecontribution of CD28 in the survival of autoreactive T cells inthe periphery, we compared the number of VSAg-reactiveTcells in the CD28(+/+) and CD28(   /   ) BALB/c mice at 3 and10 weeks, respectively. As shown in Fig. 3, CD28 deficiencyresulted in a significant increase of V b 3, 5, 11 and 12 + cells inCD4 + CD8  thymocytes, regardless of the age of the mice.Among the 3-week-old mice, the proportions of V b 11 andV b 12 + cells in the spleen were similar to those in the thymus.The V b 5-expressing cells declined in the periphery regardlessof the CD28 status, although there were still more V b 5 + cells inthe CD28(   /   ) mice. Nevertheless, for the V b 3 + CD4 Tcells,the decline was significantly faster in the CD28(   /   ) mice. By10 weeks, although CD28 deficiency continued to increasethe accumulation of autoreactive T cells in the thymus, itparadoxically reduced the number of VSAg-reactive Tcells inthe spleen. These results further support a general role forCD28 in the survival of autoreactive Tcells. CD28 deficiency resulted in accelerated cell death of autoreactive T cells in the spleen  CD28canpromoteTcellsurvivalbyinducingananti-apoptoticprotein, particularly Bcl-X L  (19). A more recent study, however,suggested that a natural CD28 ligand, CD80, induced 0001002003    P   1   A   T  g    +    P   1   C   T   L    + 82DC  +/+ 82DC  -/-    C  e   l   l  n  u  m   b  e  r   (  x   1   0    6    /   t   h  y  m  u  s   )  0.0402.002 .04000.0 a 1 10 100 1000 10000 22.4 92.1 5.6 88 7.52 1.3 6.43 6.63 76.9 4.1 2.61 8.27 110100100010000 01 24.4 1.21 5.37 gT A1P  - LTC1P  + gT A1P  + LTC1P  + 4DC    C   D   8   C   D   2   8   (   +   /   +   )   C   D   2   8   (  -   /  -   ) b 00204060800182DC  +/+ 82DC  -/-    C   D   4    +    C   D   8    +    T   h  y  m  o  c  y   t  e  s   (   %   ) 20000.0 10000.0 1000.0 1000.0 c    P   1   A   T  g   -    P   1   C   T   L    +    P   1   A   T  g    +    P   1   C   T   L    +    P   1   A   T  g   -    P   1   C   T   L    +    P   1   A   -    P   1   C   T   L    +    P   1   A    +    P   1   C   T   L    +    P   1   A   -    P   1   C   T   L    +    P   1   A    +    P   1   C   T   L    + 1101001000100001 10 100 1000 100001 10 100 1000 100001 10 100 1000 10000 Fig. 1.  CD28 deficiency attenuates the clonal deletion of transgenic mice that express both P1A antigen and P1A-reactive TCR. P1ATg + P1CTL + andP1ATg  P1CTL + micewithorwithouttargetedmutationofCD28weresacrificedat4–7weeksafterbirthandtheirthymocytenumbersofsubsetdistribution were determined by counting and flow cytometry. Data shown are the results of 2–3 independent experiments with a combined 6–8micepergroup.(a)Numbersofthymocytes.Theresultsareexpressedasmean 6 SEM. P  -valuesweredeterminedbystudent’s t  -test.(b)EffectofP1AandCD28onthethymocytesubset.Datashownasdotplotsthatdepictthedistributionofthymocytesubsets.Thepercentageofeachsubsetis marked in the quadrants. (c) Percentage of CD4 + CD8 + Tcells in the thymus of mice with different genotypes. Table 1.  Variations of clonal deletion among 4-week-oldCD28(+/   )P1CTL + P1ATg + mice I.D. Sex CD4 + CD8  CD4 + CD8 + CD4  CD8 + CD4  CD8  No. 3 10  6 1 F 4.60 13.36 21.39 60.65 12.02 F 3.90 47.96 18.54 29.58 38.03 F 2.79 53.41 18.90 24.90 52.04 M 1.95 71.31 18.18 8.56 200.05 M 2.46 81.09 8.91 7.54 130.0F1 between CD28(   /   )P1CTL + 3 CD28(+/+)P1ATg + were typed byeither PCR for P1A transgene or flow cytometry for TCR transgene.The P1CTL + P1ATg + F1 thymi were analyzed by three-color flowcytometry for subset distribution and total numbers. The numbers ofeach subset are % of gated V a 8 + thymocytes, which account for > 95% of the thymocytes. CD28 and the burden of autoreactive Tcells   1405   b  y g u e  s  t   onM a  y2  ,2  0 1  6 h  t   t   p :  /   /  i  n t  i  mm . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om  expression of Bcl-2 rather than Bcl-X L  (20). We compared theexpression of Bcl-2 by intracellular staining of the protein withPE-conjugated anti-Bcl-2 antibody. The representativeprofilesofthegatedCD8 + transgenicTcellsarepresentedinFig.4(a),and the summary data of groups of three mice are presentedin Fig. 4(b). The CD28(   /   ) P1CTL from P1ATg  miceexpressed a low level of Bcl-2. As expected, the expressionwas elevated among the T cells from P1A-transgenic mice.However, among the P1A-transgenic mice, CD28(+/   )P1CTLexpressed higher levels of Bcl-2 than the CD28(   /   )P1CTL.Consistent with differential Bcl-2 levels among CD28(+/   ) andCD28(   /   )P1CTL mice that also had P1A-transgene, we alsoobserved significantly higher proportion of cells that wereundergoing apoptosis, as judged by annexin V staining(Fig. 4c and d). Our results suggest that CD28 deficiencyresulted in increased activation-induced cell death, at least inpart by diminished induction of Bcl-2 after stimulation by themodel self antigen. In short-term adoptive transfer experiments, CD28 promoted division but not survival of P1CTL in P1A-transgenic mice  We labeled transgenic T cells from CD28(+/+) and CD28(   /   )P1CTL and transferred them into P1A-transgenic mice,and monitored the rate of division and cell death of the trans-genic Tcells. The adoptively transferred Tcells were identifiedby their expression of transgenic V a 8 as well as their ability tobind to H-L d :P1A complex as described (15,21). As shown inFig. 5(a), at 20 h after adoptive transfer, no Tcell division wasobserved, which is consistent with our previous studies usingtumor-bearing mice (15). By 42 h, CD28(+/+) T cells hadundergone 1–5 divisions. CD28 deficiency caused a sub-stantial delay in the division of Tcells, as the majority of Tcellshaveundergoneonlyonedivision.By66h,substantialdivisionof Tcells was observed in both groups, although the extent ofCFSE dilution prevented an accurate determination of thenumbers of cell divisions. To determine whether CD28deficiency also caused death of dividing Tcells, we analyzedthe transgenic Tcells for their expression of Bcl-2 and bindingto annexin V. As shown in Fig. 5(b), although a significant ca 1010010001 00001 6.52 1 10 001 0001 00001 1010010001 00001 7.52 4DC    C   D   8  A1P  - LTC1P  + )-/-(82DC gT A1P  + LTC1P  + )-/-(82DC 0 002 004 006 008 0001 1010010001 00001 71 1.11 2.62 7.54 0 002 004 006 008 0001 1010010001 00001 5.52  57.43 8.43 CSF    H   S   A gT A1P  - LTC1P  + )-/-(82DC gT A1P  + LTC1P  + )-/-(82DC b 5.46 6.75 7.51 80.1 57.7 41.1 1 10 001 0001 00001 1010010001 00001 68.3 1.72 1 10 001 0001 00001 11.4 48.3    C   D   8 gT A1P  + gT A1P  - 4DC 0 002 004 006 008 0001 002040608001 gT A1P  - gT A1P  + CSF    C  e   l   l  n  u  m   b  e  r  s d 1 10 001 0001 00001 Fig. 2.  Distinct roles for CD28 in clonal deletion and survival of autoreactive T cells. (a and b) CD8 + CD4  thymocytes from CD28(   /   )P1CTL + P1ATg + mice have similar maturity to those from the CD28(   /   )P1CTL + P1ATg  mice. Thymocytes fromtwo 7-week-old female littermateswereanalyzedbyflowcytometryusinganti-CD4,CD8andHSAantibodies.TheCD8 + CD4  cellswerefurtheranalyzedfor theirexpressionofHSAand the forward scatters. (a) Distribution of thymocyte subsets. (b) Cell size and HSA expression on the gated CD4  CD8 + thymocytes. (c) Thespleen cells from two female littermates (left panel, P1ATg  P1CTL + mice; right panel, P1ATg + P1CTL + mice) were analyzed for the presence ofCD4 and CD8 T cells in the spleen. (d) Size of the CD8 T cells in the spleen. The two groups of mice used were littermates of F1 from aCD28(   /   )P1ATg + 3 CD28(   /   )P1CTL + breeding. Table 2.  Age-dependent disappearance of CD28(   /   )autoreactive T cells in the periphery Age Sex Thymocytes 3 10  6 Spleen 3 10  6 Total CD4  CD8 + Total CD4  CD8 + 4 weeks M 210 9.87 42 4.30M 290 11.2 38 2.907 weeks M 50 11.1 22 0.38M 160 36.1 11 0.354- or 7-week-old F1 of CD28(   /   )P1CTL + 3 CD28(   /   )P1ATg + micewere analyzed for the thymocyte and spleen subsets by three-colorflow cytometry using anti-CD4, CD8 and V a 8 antibodies. Data shownare sex-matched mice that lack CD28 and express both TCR and P1Atransgenes. 1406  CD28 and the burden of autoreactive Tcells    b  y g u e  s  t   onM a  y2  ,2  0 1  6 h  t   t   p :  /   /  i  n t  i  mm . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om  proportionoftransgenicTcells underwentapoptosisat42and66 h, CD28 expression neither decreased apoptosis norincreased expression of Bcl-2. Discussion CD28 interacts with B7-1/2 and plays an important role in theinduction of T-cell immunity (22,23). Recent studies by severalgroups, including ours, have established that CD28-B7-1/2interaction plays a significant role in eliminating autoreactiveT cells in the thymus (11,12). Our new data presented in themanuscript show that mice with targeted mutation of CD28have a low number of periphery autoreactive T cells eventhough the number of autoreactive T cells in the thymus isdrastically increased.Theseresultsrevealanunexpectednewrole for CD28 in the survival of autoreactive T cells in theperiphery.At least three mechanisms can be responsible for thedisappearance of autoreactive T cells in the CD28-deficientmice. CD28(   /   ) autoreactive T cells may die immediatelyprior to their migration into the periphery. Alternatively, theautoreactive Tcells that do migrate into the periphery die afterencountering antigen, due to the lack of CD28-mediatedcostimulation. Two lines of evidence presented here supportthe alternative interpretation. First, we have observed in-creased apoptotic cells in the peripheral of TCR transgenicmice that also express the autoantigen. Second, despitecomparable numbers of autoreactive T cells in the thymus in4- and 7-week-old mice, the older mice had 5-fold lessautoreactive Tcells in the spleen. Similar conclusions can be 02468V β 3 V β 5 V β 11 V β 1202468    %    C   D   4   T  c  e   l   l  s 00.511.52OK82DC TW ** 011.53 wksSpleenThymus *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *     *      ** 10 wksV β 3 V β 5 V β 11 V β 12V β 3 V β 5 V β 11 V β 12V β 3 V β 5 V β 11 V β 120.5 Fig. 3.  Role for CD28 in both clonal deletion and survival of VSAg-reactive T cells in BALB/c mice. Thymocytes and spleen cells fromwild-type and CD28(   /   ) BALB/c mice were analyzed for CD4, CD8,and V b 3, 5, 11 and 12. Groups of three mice at 3 or 10 weeks of agewere analyzed. 020406080    M  e  a  n   f   l  u  o  r  e  s  c  e  n  c  e ba 0102030405060    A  n  n  e  x   i  n   V    +    c  e   l   l  s   % 1 10 100 1000 10000020406080100    R  e   l  a   t   i  v  e   C  e   l   l   N  u  m   b  e  r CD28 -/- P1ATg + Bcl2 expression dc P=0.02P=0.006CD28 +/- P1ATg + CD28-/-P1ATg+    R  e   l  a   t   i  v  e   C  e   l   l   N  u  m   b  e  r 0204060801001 10 100 1000 10000  Annexin V binding intensity CD28 +/- P1A + CD28 +/- P1ATg + CD28-/-P1A+ CD28 +/- P1ATg + CD28 -/- P1ATg + CD28 -/- P1ATg - CD28 +/- P1ATg + CD28 -/- P1ATg + CD28 -/- P1ATg - Fig. 4.  CD28 deficiency resulted in diminished induction of Bcl-2 (a and b) and increased apoptosis (c and d) of  ex vivo   transgenic Tcells fromP1AandP1CTLdoubletransgenicmice.(a)RepresentativeprofileofBcl-2expressionamonggatedCD8TcellsinthespleenofP1CTLtransgenicmice. (b) Means and SEM of the mean fluorescence of anti-Bcl-2 antibody staining, with three mice per group. (c) Representative histograms ofannexin V staining among gated CD8 Tcells of the spleen P1CTL transgenic mice with different genotypes. (d) Mean and SEM of % of annexin V + cells ( n  ¼ 3). The  P  -values given were calculated by student’s  t  -test. CD28 and the burden of autoreactive Tcells   1407   b  y g u e  s  t   onM a  y2  ,2  0 1  6 h  t   t   p :  /   /  i  n t  i  mm . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks