Clec9a-Targeted Antigen Evolution of B Cell Responses To

Evolution of B Cell Responses to Clec9A-Targeted Antigen Hae-Young Park, Amanda Light, Mireille H. Lahoud, Irina Caminschi, David M. Tarlinton and Ken Shortman This information is current as of September 27, 2021. J Immunol 2013; 191:4919-4925; Prepublished online 11 October 2013; doi: 10.4049/jimmunol.1301947 http://www.jimmunol.org/content/191/10/4919 Downloaded from

References This article cites 20 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/191/10/4919.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 27, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Evolution of B Cell Responses to Clec9A-Targeted Antigen

Hae-Young Park,*,† Amanda Light,* Mireille H. Lahoud,†,‡ Irina Caminschi,†,x David M. Tarlinton,*,{ and Ken Shortman*,†,{

The response of B cells to Ag targeted to Clec9A on dendritic cells was followed using the hapten nitrophenol (NP) conjugated to rat Ig carrier. Injection of small amounts of NP conjugated to anti-Clec9A in the absence of adjuvants gave high and very prolonged Ab responses, approaching those obtained by high doses of nontargeted NP–protein conjugates with alum adjuvant. The response to NP–anti-Clec9A included the transient formation of germinal centers, maturation of Ab afﬁnity, and some memory B cell formation. Serum Ab titers remained high 35 wk postimmunization, well after the initial follicular response had faded. The results suggest Clec9A-targeting strategies for improving Ab responses to vaccine Ags. The Journal of Immunology, 2013, 191: 4919–4925.

argeting Ag to dendritic cells (DCs), by injecting Ag is the high Ab responses obtained even in the absence of adjuvants

coupled to a mAb specific for a DC surface molecule, is an or signs of DC activation (9). One reason for this is a prolonged Downloaded from T established experimental strategy for enhancing immune CD4 T cell response that results in the production of T follicular responses and has the potential for improving the efficiency of helper cells (TFH), which in turn is a result of persistence of the vaccines (1–3). Such enhanced immune responses are considered Ag–anti-Clec9A construct in the serum and consequent prolonged to result from more effective presentation of the Ag by the tar- Ag presentation by the targeted DCs (11). This result differs from geted DC to Ag-specific T cells. The C-type lectin-like molecule targeting other DC surface molecules such as DEC-205. With DEC-

Clec9A (also called DNGR1) has proved to be a very effective 205, the targeting construct is rapidly removed from the serum, the http://www.jimmunol.org/ target molecule for this application. Clec9A is a dead cell rec- T cell response is transient, regulatory T cells are generated, and Ab ognition molecule, the ligand being F-actin exposed when the cell responses are minimal unless the DCs are activated with adjuvants membrane is disrupted (4–6). An advantage of Clec9A for tar- (11, 14). However, other observations have been made of effective geting applications is that expression is restricted to certain DCs. Ab responses on targeting Ag to other DC surface molecules, even In mice, Clec9A is selectively expressed by the CD8+ DC lineage in the absence of adjuvant (3). In one recent study targeting nitro- and its migratory equivalent, the CD103+ DCs, with lower ex- phenol (NP)–protein conjugates to DC-inhibitory receptor 2 on the 2 pression on plasmacytoid DCs; it is not expressed on lymphocytes surface of a subset of marginal zone CD8 DCs, an effective Ab or other hematopoietic cells (7–10). In humans it is expressed on response to NP was obtained in the absence of adjuvant (15). This BDCA3+ DCs, the equivalent of the murine CD8+ DC lineage. finding was shown to be an extrafollicular but T cell–dependent by guest on September 27, 2021 Targeting Ag to Clec9A results in very effective CD4 and CD8 response, strongly enhanced by the DC-induced helper T cell gen- T cell proliferative responses; with coadministration of adjuvants eration, dependent on an indirect pickup and subsequent presenta- to activate the DCs, this leads to the generation of effector cyto- tion of Ag by the B cell component. toxic T cells (8, 9, 11). In this latter respect, targeting Clec9A We now focus on the B cell response when Ag is targeted to gives responses similar to those obtained by targeting DEC-205, Clec9A, using the well-established B cell hapten NP (16) conju- another molecule strongly expressed by CD8+ DCs (11–13). gated to a mAb against Clec9A, which, being a rat Ig, serves as One surprising feature of the responses obtained by targeting carrier with an effective helper T cell epitope in C57BL/6 mice protein Ag to DCs, using the anti-Clec9A mAb we have generated, (9). Targeting tiny amounts of NP-Ig to Clec9A in the absence of adjuvant gave very high anti-NP Ab responses with the initial features of a follicular response, including some affinity matura- *The Walter and Eliza Hall Institute, Parkville, Victoria 3052, Australia; †Burnet tion and memory B cell generation. The serum response was pro- Institute, Melbourne, Victoria 3004, Australia; ‡Department of Immunology, Monash longed, Ab persisting at high amounts even when other aspects of x University, Melbourne, Victoria 3004, Australia; Department of Microbiology and the B cell response had diminished. Because protection against Immunology, University of Melbourne, Melbourne, Victoria 3010, Australia; and {Department of Medical Biology, University of Melbourne, Melbourne, Victoria many infectious diseases depends on Ab production, these findings 3010, Australia have many implications for the enhancement of vaccine effective- Received for publication July 22, 2013. Accepted for publication September 12, ness by targeting to Clec9A. 2013. This work was supported by the National Health and Medical Research Council, Materials and Methods Australia; the Victorian State Government Operational Infrastructure Support; and the National Health and Medical Research Council, Australia, Independent Research Animals Institutes Infrastructure Support Scheme. Themiceusedwere8-to12-wk-oldC57BL/6females,bredandmain- Address correspondence and reprint requests to Prof. Ken Shortman or Dr. Hae- tained under specific pathogen–free conditions at the Walter and Eliza Young Park, The Walter and Eliza Hall Institute, Immunology Division, 1G Royal Hall Institute. All animal procedures were approved by the Walter and Eliza Parade, Parkville, VIC 3050, Australia (K.S.) or Burnet Institute, Centre for Biomed- Hall Institute Animal Ethics Committee. icine, 85 Commercial Road, Melbourne, VIC 3004, Australia (H.-Y.P.). E-mail ad- dresses: [email protected] (K.S.) and [email protected] (H.-Y.P.) Ags and immunizations Abbreviations used in this article: AFC, Ab-forming cell; BM, bone marrow; DC, dendritic cell; GC, germinal center; KLH, keyhole limpet hemocyanin; NP, 4(hy- Rat anti-Clec9A mAb (clone 24/04-10B4 9.5) or an isotype control rat droxyl-3-nitrophenyl)acetyl; TFH, T follicular helper cell. IgG2a mAb (clone GL117) (9) was conjugated to the hapten 4(hydroxyl-3- nitrophenyl)acetyl, as previously described (16). Final NP to Ig conjuga- Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 tion ratios for the NP–anti-Clec9A ranged from NP7 to NP8, as determined www.jimmunol.org/cgi/doi/10.4049/jimmunol.1301947 4920 B CELL RESPONSES TO Clec9A-TARGETED Ag

by spectrophotometry. The NP coupled to keyhole limpet hemocyanin calculated as the ratio of the number of spots in the NP1-BSA–coated wells (KLH) (NP19–NP20) was precipitated in alum (16, 17). Immunizations to the number of spots in the NP16-BSA–coated wells. were usually performed using 2 mg NP-conjugated Abs (anti-Clec9A–NP and isotype control–NP) administered by i.v. injection (9) or with 100 mg Transfer assay for memory B cells alum-precipitated NP-KLH (NP-KLH/alum) administered by i.p. injection Mice were immunized with 2 mg anti-Clec9A–NP or isotype control–NP. (16). After 42 d, B cells were isolated from each group of immunized mice and ELISA assay for anti-NP serum Ab from unimmunized mice using a B Cell Isolation Kit (130-090-862; Mil- tenyi Biotec). The isolated B cells (purity of B220+ B cells was . 98%) Anti-NP Ab titers were normally determined by ELISA assay exactly as were injected i.v. into recipient mice (15–18 3 106 cells per mouse) that described (16). To calculate the exact titer and the anti-NP IgG1 concen- had been preimmunized with 50 mg KLH/alum 42 d previously. The re- tration, a standard curve was made using GSX2.1, an IgG1 NP-specific cipient mice were then injected i.p. with 50 mg soluble NP-KLH without 5 hybridoma utilizing VH186.2 (18). On the basis of this curve, a titer of 10 alum. Blood was sampled at the indicated times before and after transfer corresponded to 388 mg/ml IgG1, whereas a titer of 104 corresponded to of B cells and Ag injection. 24 mg/ml. For the other Ig isotypes for which standard curves were not constructed, an endpoint titer of 10 SDs above a no serum-negative control Immunohistology was used. Spleens from the immunized mice were embedded and frozen in OCT ELISA assay for anti-rat Ig serum Ab (Tissue Tek; Sakura) at 280˚C. Sections of size 7 mM were cut using a microtome (Leica) and then stained as follows: rat anti-B220 (RA3- For anti-rat Ig serum levels, the ELISA assay was as previously described 6B2) and biotinylated peanut agglutinin, then anti-rat k-HRPO (3090- (9). In the blocking assay for relative affinity, serum was first preincubated 05; Southern Biotech) and streptavidin alkaline phosphatase (7100-04; with 10 mgor50mg rat Ig for 30 min at room temperature, and the mixture Southern Biotech). The color was developed using AEC Substrate was then transferred into the ELISA assay plates. The percentage blocking (SK4200; Vector Laboratories) and FastBlue (SK5300; Vector Labora- Downloaded from by rat Ig was enumerated based on the OD readings given by unblocked tories) for visualization of the peroxidase and the alkaline phosphatase, serum samples. respectively. Images were taken on an Axioplan 2 imaging microscope (Carl Zeiss). NP-specific B cell assays Statistics Ag-specific B cells were prepared and stained for flow cytometric analysis exactly as described (16). The Ag-specific B cells were identified as IgM2 Statistical analysis was performed using Prism software (GraphPad 2 2 + + +

IgD Gr-1 B220 IgG1 NP cells. Data were analyzed using LSRII (BD) software). A two-tailed unpaired Student t test was used for differences http://www.jimmunol.org/ and Weasel software (Walter and Eliza Hall Institute). ELISPOT assays between anti-Clec9A NP and isotype control NP, and one-way ANOVA were as previously published (16). For enumerating Ab-forming cells (AFCs), was used for differences among three groups. The p values are indi- spots were counted in the NP16-BSA–coated well and divided by the cated as follows: *p , 0.05, **p , 0.01, ***p , 0.001, and ****p , number of seeded total cells per well. The relative afﬁnity maturation was 0.0001. by guest on September 27, 2021

FIGURE 1. Targeting NP to Clec9A generates AFCs and long-lived Ab responses. Groups of C57BL/6 mice were immunized i.v. with 2 mg NP- conjugated rat anti-Clec9A (mAb 24/04-10B4), or 2 mg NP-conjugated isotype control rat anti-IgG2a (mAb GL117) or, as a high response control, 100 mg NP-KLH in alum. (A) The titer of anti-NP IgG1 Ab, the dominant Ig isotype, was determined by ELISA at the indicated times postimmunization using 5 4 NP16 BSA for Ab capture. A titer of 10 corresponded to 388 mg/ml, and a titer of 10 corresponded to 24 mg/ml of a standard anti-NP IgG1. (B) The proportion of NP-specific IgG1+ B cells from the spleen was enumerated by flow cytometry. The frequency of NP-specific IgG1 AFCs per 106 cells among (C) splenocytes or (D) BM cells was determined by ELISPOT. Data are pooled from four (A) or two (B–D) experiments with 4–10 mice per time point, and show the mean 6 SEM. Data were analyzed by an unpaired two-tailed Student t test, and significance was indicated as *p , 0.05, **p , 0.01, ***p , 0.001. The Journal of Immunology 4921

Results to Clec9A in the absence of adjuvants gave Ab responses com- Serum Ab responses on targeting NP to Clec9A parable to those obtained using adjuvants (9, 11). However, for the To determine if the high Ab responses previously obtained by same small amount of NP-protein injected, targeting to Clec9A targeting protein Ag to Clec9A would also be obtained to a gave a better response even at the peak than that obtained using deﬁned B cell hapten, 2 mg of NP conjugated to anti-Clec9A alum as adjuvant. Thus, in one experiment, when 2 mg of NP- mAb (the rat IgG2a mAb 24/04-10B4) (9) was injected i.v. into isotype control construct instead of 100 mg of NP-KLH was ad- C57BL/6 mice. The response was compared with an i.v. injec- ministered i.p. with alum adjuvant, only 0.003 6 0.002 mg/ml of tion of 2 mg of NP conjugated to an isotype-matched non- anti-NP IgG1 was obtained by 2 wk, whereas 2 mg of NP–anti- targeting mAb (the rat IgG2a mAb GL117). The response to Clec9A injected i.v. without adjuvant gave 0.24 6 0.07 mg/ml of these small amounts of targeted NP without adjuvant was com- anti-NP IgG1. pared with the response obtained under conditions known to give B cell responses on targeting NP to Clec9A high Ab titers, namely, injection of 100 mg of NP coupled to KLH in alum as adjuvant. Targeting NP–rat Ig to Clec9A in this way An advantage of using NP as Ag is that techniques are available gave a strong serum Ab response with a single injection, com- for monitoring the dynamics of the response of the NP-speciﬁc paredwiththemarginalresponseobtained following injection B-lineage cells. These techniques were used to examine the basis of untargeted NP–rat Ig (Fig. 1A). The response peaked 2–3 wk of the serum Ab responses. Following injection of the NP–anti- post injection with an anti–NP IgG1 titer of corresponding to Clec9A complex, NP-binding B cells expanded to represent up to 0.39 mg/ml. The response dropped after 3 wk, declining slowly

3% of the total spleen B cells at day 7, although this number Downloaded from such that significant serum Ab titers were maintained beyond dropped to much lower levels by 2 wk and remained low thereafter 35 wk after this single injection. Although IgG1 dominated the (Fig. 1B). No such response was detected with the same quantity anti-NP Ab response, other Ig isotypes were also produced (Fig. of untargeted NP-protein. The response was below that obtained 2). Overall, the targeted anti-NP Ab response was similar to those using a high dose of NP-KLH with alum, in line with the initial previously reported using the protein Ags OVA and rat Ig, which difference in Ab production. The frequency of NP-specific AFCs were shown to be T cell and DC dependent (9, 11). Although the anti-NP Ab response obtained by injecting 2 mgof in the spleen, as detected by ELISPOT also peaked around day 7, http://www.jimmunol.org/ NP–anti-Clec9A without adjuvant was at the peak a little less than and the number was only a little lower than that obtained with that obtained by i.p. injection of 100 mg of NP-KLH in the high-dose NP-KLH with alum (Fig. 1C). However, the frequency presence of alum as adjuvant, the persistence of the response was of NP-specific AFC following a low dose of targeted NP then fell more marked, so by 4 wk the titers were equivalent (Fig. 1A). This rapidly after day 7, much more so than was observed following the finding is in line with the results for protein Ag, in which targeting high-dose NP-KLH with alum. by guest on September 27, 2021

FIGURE 2. Targeting NP to Clec9A induces generation of anti-NP Ab subtypes in addition to IgG1. Groups of C57BL/6 mice were immunized as in Fig. 1, and generation of other serum Ig subtypes, determined by ELISA, were as follows: (A) IgG2b (n =5– 8), (B) IgG2c (n = 4–5), (C) IgG3 (n = 4–8), (D) IgE (n = 3–5), and (E) IgM (n = 3–5). Titers are shown as mean 6 SEM. In contrast to Fig. 1, standards were not available to compare titers with absolute serum concentrations, so titers are not directly comparable. Data were analyzed by an unpaired two-tailed Student t test, and signiﬁcance was indicated as *p , 0.05, **p , 0.01. 4922 B CELL RESPONSES TO Clec9A-TARGETED Ag

The results in Fig. 1 revealed a surprising characteristic of the Germinal center formation and affinity maturation on targeting response to the Clec9A-targeted Ag. In the case of high-dose NP to Clec9A NP-KLH in alum, the frequency of NP-specific AFCs detected Our previous studies on the Clec9A-targeted response to OVA had in the spleen over the 6-wk period was reasonably in line with demonstrated a remarkable generation of TFH cells, some of which the anti-NP serum Ab titer. However, in the case of the low-dose were located in the GCs of the B cell follicles (11). Accordingly, Clec9A-targeted Ag, the frequency of NP-specific AFCs in the we asked whether the B cell response to Clec9A-targeted NP 3 spleen dropped around 10 -fold, whereas the NP-specific Ab would involve germinal center (GC) formation and affinity mat- titer dropped only 10-fold. The “missing” AFCs were sought uration of the anti-NP Ab. Affinity maturation at the level of the elsewhere in the mice, particularly in bone marrow (BM). In the individual AFC was measured by comparing the ELISPOT count case of high-dose NP-KLH plus alum, AFCs in the BM in- using an Ab capture protein with 1 NP conjugated per protein creased between 1 and 3 wk post injection, but in the case of molecule (high affinity) versus 16 NP per molecule (total affinity). the low-dose Clec9A-targeted Ag, the AFC frequency in BM Following targeting of NP to Clec9A, an increase in the relative dropped over this period (Fig. 1D). Other sites in the Clec9A- number of high affinity AFCs was observed over the initial 5- to targeted mice, including blood, lymph nodes, and GALT, did not 14-d period, an increase similar to that seen upon immunization reveal any AFC accumulation or high AFC frequency over this with NP-KLH in alum (Fig. 3A). Although accurate assessment period. However, it is important to note that 6 wk after injection could not be made beyond this time because of the low number of the Clec9A-targeted mice retained AFCs in spleen and BM at AFCs present, it was clear that the relative affinity of the AFCs a frequency above that detected following injection of the same produced following Clec9A targeting fell after day 14, indicating quantity of untargeted NP-rat Ig (Fig. 1C, 1D). We concluded that the higher affinity AFCs had disappeared from the spleen Downloaded from that the persistent high serum Ab titers following targeting of (Fig. 3A). In contrast, continuing production of high affinity AFCs NP-rat Ig to Clec9A were produced by a relatively low frequency was maintained in the spleen following immunization with NP- of persistent, efficient AFCs. KLH/alum. In contrast to the ELISPOT data, the relative affinity http://www.jimmunol.org/

FIGURE 3. Targeting NP to Clec9A involves early GC formation and results in affinity maturation. Groups of C57BL/6 mice were immunized as in Fig. 1. (A) Affinity maturation at the AFC level was by guest on September 27, 2021 determined by ELISPOT as the ratio of high affinity (NP1 binding) to total affinity (NP16 binding) IgG1 NP-specific AFCs at days 5 (n = 8), 7 (n = 5–6), and 14 (n = 6–12) postimmunization. Results are the mean 6 SEM. (B) Affinity maturation at the level of serum IgG1 Ab was measured by ELISA as the relative titer, using as capture Ag NP1 BSA (high affinity) versus NP16 BSA (total affinity). Results represent two experiments with five mice per time point. The error bar shows mean 6 SEM. As in Fig. 2, a thicker lined circle, a gray marked circle, and a thinner lined circle represents mice that were immunized with anti-Clec9A–NP, isotype control–NP, and NP-KLH/alum, respectively. (C) GC staining of spleens from C57BL/6 mice 11 d postimmunization. GCs were revealed with peanut agglutinin (blue) and B cell follicles with anti-B220 (red). Scale bars, 100 mm. The histology shown is representative of six mice from two experiments. The graph shows the percentage of follicles with GCs as the mean 6 SEM. Data were analyzed by an unpaired two-tailed Stu- dent t test and significance was indicated as *p , 0.05, **p , 0.01, ***p , 0.001. The Journal of Immunology 4923 of the serum Ab produced in the mice could be accurately mon- itored over the entire response, using the relative titer when NP1 versus NP16 was used as the capture Ag in the ELISA assay (Fig. 3B). In response to Clec9A-targeted NP, there was an increase in the affinity of the anti-NP Ab produced over the first 28 d and especially over the first 14 d. However, in accordance with the transient production of high affinity AFCs, serum Ab affinity maturation was substantially less than that obtained with the high-dose NP-KLH/alum schedule, in which high affinity AFCs persisted in the spleen. The incidence of GCs in the spleen B cell follicles following immunization with NP–anti-Clec9A was assessed by immuno- histochemistry. An increase in GCs, compared with GCs in mice immunized with NP conjugated to the nontargeting isotype-matched control,wasseenby7dandwasstatisticallysignificantat11d(Fig. 3C). This difference disappeared by 14 d. In contrast, immunization with NP-KLH and alum gave a higher incidence of follicles, with GCs maintained over the entire 1- to 2-wk period (data not shown).

Overall, we concluded that part of the early response to targeting Downloaded from tiny amounts of NP–rat Ig to Clec9A involves a follicular response, with GC formation and Ab affinity maturation. However, this GC response is transient and shorter than that obtained with the NP-KLH/alum schedule. Alignment of the anti-rat Ig response with the anti-NP response http://www.jimmunol.org/ Because our previous studies had been with protein Ag, it was important to know if the anti-NP Ab response resembled the anti- protein Ab response. Because the Clec9A-targeting mAb and the carrier protein for the NP immunization was a rat IgG2a, it was possible to monitor the serum anti-rat Ig response as well as the anti-NP response to the same immunogen. The anti-rat Ig titer (Fig. FIGURE 4. Targeting NP-rat Ig to Clec9A effectively induces an anti-rat 4A) showed a similar time course to that of the anti-NP titer (Fig. Ig response as well as an anti-NP response. Groups of C57BL/6 mice were 1A), with a similarly prolonged Ab response. We had no assay for immunized as in Fig. 1. (A) Serum anti-rat Ig reactivity was measured by affinity maturation of the anti-rat Ig response at the AFC level, but ELISA at the indicated time points. Results represent pooled data from two by guest on September 27, 2021 an estimate of relative affinity at the serum Ab level could be experiments with at least five mice per time point, and are shown as the obtained by measuring the extent of blocking of the final ELISA mean 6 SEM. (B) The relative affinity of the serum was determined as the assay by preincubation of the serum dilutions with various con- percent blocking of the ELISA readings by preincubation with various centrations of rat Ig. Over the 7- to 42-d period, the anti-rat Ig in kinds of concentration of rat Ig. Data are pooled from two experiments with six to eight mice per time point, and expressed as the mean 6 SEM. the serum became more readily blocked by preincubation with rat Data were analyzed by an unpaired two-tailed Student t test and signifi- Ig, so accordingly increased in affinity (Fig. 4B). By these criteria cance was indicated as *p , 0.05. the anti-NP response appeared similar to the anti-protein response.

Are memory B cells generated on targeting NP to Clec9A? with nontargeting NP–isotype-matched control, or from mice We had previously demonstrated that the initial Ab response to immunized with 100 mg of NP-KLH and alum, or from unim- protein Ag targeted to Clec9A could be boosted, once the serum munized mice. The B cells were then transferred into recipient Ab had declined from the peak, by a secondary injection of even mice that had been preimmunized with KLH in alum to produce untargeted free Ag [(9); I. Caminschi and K. Shortman, unpub- KLH-specific Th cells. These recipient mice were then challenged lished observations]. However, whether this indicated the presence by injection of NP-KLH, and the total affinity anti-NP Ab re- of memory B or memory T cells, or was due to other effects, was sponse followed (Fig. 5). When the total serum anti-NP response not clear. We tested the effect of secondary injections of either 2 was measured, little sign of memory B cell generation was pres- mg of targeted NP–rat Ig, or 50 mg of untargeted NP–rat Ig, at 14 ent, not even from mice primed with the high-dose NP-KLH plus and 21 d after the primary targeted immunization but observed no alum schedule. However, when the readout was restricted to high boost to the already high anti-NP titers. A significant boost to the affinity serum Ab, it was clear that the B cells from mice pre- anti-NP titer was obtained in one experiment by the second in- immunized by Clec9A targeting and by high Ag dose plus adju- jection of 2 mg of targeted NP-rat Ig 242 d after the first injection; vant gave a more rapid response with a higher titer at the early, this appeared to be a true memory response because the Ab 5- to 7-d period (Fig. 5). This difference in titer was no longer ev- produced showed a significantly increased affinity. However, we ident over the day 14–28 peak response period. We concluded that sought more direct evidence for memory at the B cell level. Clec9A targeting of NP did generate some rapidly responding, high The rapid decline in NP-specific B cells when NP was targeted affinity memory B cells, but their contribution was eventually to Clec9A (Fig. 1B) argued against the presence of a very large swamped by the large response from the primary B cells. memory B cell generation relative to the primary B cell response. To determine in a functional assay if any memory B cells were Discussion generated, B cells were isolated from mice 42 d postimmunization This study, using the well-characterized B cell epitope NP, has with 2 mg of NP–anti-Clec9A, or from mice similarly immunized confirmed and extended the main findings from our previous work 4924 B CELL RESPONSES TO Clec9A-TARGETED Ag

FIGURE 5. Targeting NP to Clec9A generates some high affinity memory B cells. Groups of C57BL/6 mice were immunized as in Fig. 1. After 42 d, B220+ B cells were isolated from each immunized mice group as well as from an unimmunized control group. The B cells were transferred into recipient C57BL/6 mice that had been immunized with 50 mg KLH in alum. The recipient mice were then challenged by injection of 50 mg soluble NP-KLH without alum. The titer of serum anti-NP IgG1 Ab was determined by ELISA at the indicated times after transfer, using NP1 BSA (high affinity) or NP16 BSA (total affinity) for Ab capture. Titers are shown as mean 6 SEM. Results represent one experiment with five mice per time point; the results were confirmed in a second experiment. Data were analyzed by one-way ANOVA analysis, followed Downloaded from by multiple comparisons, with significance indicated as ****p , 0.0001. http://www.jimmunol.org/ with protein Ags (9, 11), namely, that targeting small amounts higher affinity Abs, either more Abs or Abs with prolonged se- of Ag to Clec9A on DCs using an anti-Clec9A mAb produces rum persistence, compared with the AFCs seen early in the re- a strong and persistent Ab response even without adjuvants. The sponse. They also appear more efficient than the AFCs remaining initial response to Clec9A-targeted NP has the characteristics of after the conventional immunization with a high dose of Ag in a follicular response, with a transient formation of GCs, a degree alum adjuvant. of affinity maturation of the Ab produced, and the generation of Effective immunization should involve the generation of memory some fast-responding high affinity memory B cells. The likely B cells. Although the response to protein Ag can be boosted by by guest on September 27, 2021 mechanism would be that the NP–rat Ig targeted to Clec9A on the secondary injection of the Ag, we obtained little boosting of the Ab surface of DCs of the CD8+ DC family leads to effective Ag response to NP; this may have been a consequence of the primary presentation by these DCs. The persistence of anti-Clec9A in the response being so high, at a near maximal ceiling. The very high serum, as previously demonstrated (11), leads to continuous load- primary response to NP also made it difficult to detect B memory ing of rat Ig T cell epitopes on MHC class II, even in the absence cell generation, even with the high-dose NP-KLH plus adjuvant of DC activation. In line with the striking generation of TFH cells system. Despite this swamping by a primary response, it was possible found using OVA as Ag (11), these conditions should promote to detect some rapidly responding, high affinity NP-specific memory efficient production of TFH cells specific for rat Ig. In the simplest B cell activity in mice primed by targeting NP to Clec9A. model, the persistence of NP–anti-Clec9A in the serum also allows The high and persisting Ab responses arising from a single B cells to encounter the NP hapten, and the TFH/B cell interaction injection of tiny amounts of Ag targeted to Clec9A, together with then drives the follicular response. the finding that this response can be obtained without the sec- One unresolved issue with this model is whether the enhanced ondary effects of adjuvants, suggests this approach could be applied Ab response on targeting Ag to Clec9A on DCs can be entirely to putative vaccine Ags for certain infectious diseases in which Ab explained by the efficient production of Th cells, specifically TFH is known to provide protection. Our experiments with NP as the cells, that we have previously documented (11). Another possi- model Ag provide some information relevant to this objective. bility is that the Clec9A+ DCs may present Ag to both T and Clearly, for Ab responses the Ag–anti-Clec9A construct used for B cells. Several studies have suggested that DCs may present Ag targeting should maintain the B cell epitopes intact—not always to B as well as to T cells (19, 20). a simple task if conformational B cell epitopes are involved. An One striking feature of the Ab response to NP following tar- alternative is not only to target the T cell epitopes to Clec9A to geting to Clec9A not readily explained by these models is the produce enhanced T cell help but also to use some form of prolonged serum Ab titer, persisting well after signs of a GC re- untargeted Ag containing the B cell epitopes to induce B cell sponse have faded and when the number of AFCs in the spleen, activation. BM, and other sites has dropped to a relatively low level. A Model Ag used for laboratory studies has been selected as giv- proportion of the targeted response may be extrafollicular, as has ing good immune responses, probably reflecting a naturally high been reported for the anti-NP response on targeting DC-inhibitory frequency of precursors responsive to that Ag. The NP system may receptor 2 on DCs (15). According to this model, the DCs may be an extreme example of this situation, because the primary re- “hand on” the Ag to B cells, which then provide ongoing Ag sponse is so high it was difficult to see a secondary boost. Such an presentation to T cells. Regardless of the mechanism driving the initially high incidence of Ag-specific B or T cells may not apply prolonged Ab response, the result suggests that the few remaining to many putative vaccine Ags, for which a low initial precursor AFCs are very efficient, producing, in addition to the observed frequency means even a primary response enhanced by Clec9A The Journal of Immunology 4925 targeting may be inadequate for protective immunity. In such cases 4. Sancho, D., O. P. Joffre, A. M. Keller, N. C. Rogers, D. Martıńez, P. Hernanz- Falcoń, I. Rosewell, and C. Reis e Sousa. 2009. Identification of a dendritic cell some form of priming, possibly requiring adjuvants, may be needed receptor that couples sensing of necrosis to immunity. Nature 458: 899–903. to enhance the precursor frequency prior to a boost involving 5. Zhang, J. G., P. E. Czabotar, A. N. Policheni, I. Caminschi, S. S. Wan, Clec9A targeting. S. Kitsoulis, K. M. Tullett, A. Y. Robin, R. Brammananth, M. F. van Delft, et al. 2012. The dendritic cell receptor Clec9A binds damaged cells via exposed actin Our experiments point to some issues that need to be overcome filaments. Immunity 36: 646–657. prior to such a prime-boost approach. One problem with a sec- 6. Ahrens, S., S. Zelenay, D. Sancho, P. Hanc, S. Kjær, C. Feest, G. Fletcher, ondary injection of Clec9A-targeted Ag is that any Ab generated in C. Durkin, A. Postigo, M. Skehel, et al. 2012. F-actin is an evolutionarily conserved damage-associated molecular pattern recognized by DNGR-1, a re- the primary injection would remove the targeting construct from ceptor for dead cells. Immunity 36: 635–645. the bloodstream; we have demonstrated that prolonged persistence 7. Huysamen, C., J. A. Willment, K. M. Dennehy, and G. D. Brown. 2008. CLEC9A is a novel activation C-type lectin-like receptor expressed on BDCA3+ of the Ag–anti-Clec9A construct in the bloodstream is a major dendritic cells and a subset of monocytes. J. Biol. Chem. 283: 16693–16701. reason for its effectiveness (11). Generation of memory B and 8. Sancho, D., D. Mouraõ-Sa´, O. P. Joffre, O. Schulz, N. C. Rogers, D. J. Pennington, T cells without generating a blocking Ab is a requirement. This J. R. Carlyle, and C. Reis e Sousa. 2008. Tumor therapy in mice via antigen targeting to a novel, DC-restricted C-type lectin. J. Clin. Invest. 118: 2098–2110. limitation may be overcome using a different form of the Ag for 9. Caminschi, I., A. I. Proietto, F. Ahmet, S. Kitsoulis, J. Shin Teh, J. C. Lo, the prime or the boost, limiting the targeting construct to a single A. Rizzitelli, L. Wu, D. Vremec, S. L. van Dommelen, et al. 2008. The dendritic administration. cell subtype-restricted C-type lectin Clec9A is a target for vaccine enhancement. Blood 112: 3264–3273. These likely problems do not seem insurmountable, and many 10. Desch, A. N., G. J. Randolph, K. Murphy, E. L. Gautier, R. M. Kedl, solutions could emerge from a detailed study of the Clec9A- M. H. Lahoud, I. Caminschi, K. Shortman, P. M. Henson, and C. V. Jakubzick. 2011. CD103+ pulmonary dendritic cells preferentially acquire and present targeted response, even using laboratory model Ags. It is impor- apoptotic cell-associated antigen. J. Exp. Med. 208: 1789–1797. tant that CLEC9A is expressed in a similar way in the human DC 11. Lahoud, M. H., F. Ahmet, S. Kitsoulis, S. S. Wan, D. Vremec, C. N. Lee, Downloaded from system (7–9), and we have found that mAb against human B. Phipson, W. Shi, G. K. Smyth, A. M. Lew, et al. 2011. Targeting antigen to mouse dendritic cells via Clec9A induces potent CD4 T cell responses biased CLEC9A can serve to target protein Ags in primates and enhance toward a follicular helper phenotype. J. Immunol. 187: 842–850. Ab production in a way that closely parallels the effects in mice 12. Bonifaz, L. C., D. P. Bonnyay, A. Charalambous, D. I. Darguste, S. Fujii, (M.H. Lahoud, R. De Rose, S.J. Kent, A.M. Lew, K. Shortman, H. Soares, M. K. Brimnes, B. Moltedo, T. M. Moran, and R. M. Steinman. 2004. In vivo targeting of antigens to maturing dendritic cells via the DEC-205 re- and I. Caminschi, manuscript in preparation). The use of CLEC9A ceptor improves T cell vaccination. J. Exp. Med. 199: 815–824. targeting to enhance immunity and improve vaccines therefore has 13. Idoyaga, J., A. Lubkin, C. Fiorese, M. H. Lahoud, I. Caminschi, Y. Huang, http://www.jimmunol.org/ A. Rodriguez, B. E. Clausen, C. G. Park, C. Trumpfheller, and R. M. Steinman. sufficient promise to justify further detailed study of the immu- 2011. Comparable T helper 1 (Th1) and CD8 T-cell immunity by targeting HIV nological mechanisms involved and to devise strategies for its gag p24 to CD8 dendritic cells within antibodies to Langerin, DEC205, and effective implementation. Clec9A. Proc. Natl. Acad. Sci. USA 108: 2384–2389. 14. Hawiger, D., K. Inaba, Y. Dorsett, M. Guo, K. Mahnke, M. Rivera, J. V. Ravetch, R. M. Steinman, and M. C. Nussenzweig. 2001. Dendritic cells induce peripheral Acknowledgments T cell unresponsiveness under steady state conditions in vivo. J. Exp. Med. 194: 769–779. We thank Prof. William R. Heath (University of Melbourne, Melbourne, 15. Chappell, C. P., K. E. Draves, N. V. Giltiay, and E. A. Clark. 2012. Extra- Australia) for helpful discussions and for critical reading of this manuscript. follicular B cell activation by marginal zone dendritic cells drives T cell- dependent antibody responses. J. Exp. Med. 209: 1825–1840.

16. Zotos, D., J. M. Coquet, Y. Zhang, A. Light, K. D’Costa, A. Kallies, by guest on September 27, 2021 Disclosures L. M. Corcoran, D. I. Godfrey, K. M. Toellner, M. J. Smyth, et al. 2010. IL-21 The authors have no ﬁnancial conﬂicts of interest. regulates germinal center B cell differentiation and proliferation through a B cell- intrinsic mechanism. J. Exp. Med. 207: 365–378. 17. Blink, E. J., A. Light, A. Kallies, S. L. Nutt, P. D. Hodgkin, and D. M. Tarlinton. 2005. Early appearance of germinal center-derived memory B cells and plasma References cells in blood after primary immunization. J. Exp. Med. 201: 545–554. 1. Tacken, P. J., I. J. de Vries, R. Torensma, and C. G. Figdor. 2007. Dendritic-cell 18. McHeyzer-Williams, M. G., G. J. Nossal, and P. A. Lalor. 1991. Molecular immunotherapy: from ex vivo loading to in vivo targeting. Nat. Rev. Immunol. 7: characterization of single memory B cells. Nature 350: 502–505. 790–802. 19. Wykes, M., A. Pombo, C. Jenkins, and G. G. MacPherson. 1998. Dendritic cells 2. Trumpfheller, C., M. P. Longhi, M. Caskey, J. Idoyaga, L. Bozzacco, T. Keler, interact directly with naive B lymphocytes to transfer antigen and initiate class S. J. Schlesinger, and R. M. Steinman. 2012. Dendritic cell-targeted protein vac- switching in a primary T-dependent response. J. Immunol. 161: 1313–1319. cines: a novel approach to induce T-cell immunity. J. Intern. Med. 271: 183–192. 20. Qi, H., J. G. Egen, A. Y. Huang, and R. N. Germain. 2006. Extrafollicular ac- 3. Caminschi, I., and K. Shortman. 2012. Boosting antibody responses by targeting tivation of lymph node B cells by antigen-bearing dendritic cells. Science 312: antigens to dendritic cells. Trends Immunol. 33: 71–77. 1672–1676.