Connect with us

Latest Posts

Modular cytokine receptor-targeting chimeras for targeted degradation of cell surface and extracellular proteins




Overexpression of PD-L1 on cancer cells leads to engagement of T cell checkpoint protein PD-1 and suppression of cytotoxic T cell activity21. To demonstrate proof of concept, we generated a KineTAC targeting PD-L1 using generic knob-in-hole bispecific antibody–cytokine fusions22, in which the human CXCL12 chemokine was N-terminally fused to the knob Fc domain, and the second arm contained the antigen-binding fragment (Fab) antibody sequence for atezolizumab (Atz; Tecentriq), a Food and Drug Administration (FDA)-approved inhibitor of PD-L1, fused to the hole Fc domain (Fig. 1a). The T336W mutation and T366S, L368A and Y407V mutations were introduced into the IgG1 Fc domain to generate the knob and hole Fc domains, respectively. For mechanistic simplification, the N297G mutation was also introduced to remove effector functions.

One advantage of KineTACs is that production of cytokine-bearing bispecifics is not complicated by light chain–heavy chain mispairing problems, which are common to bispecific IgGs containing Fabs on both arms. Moreover, the KineTAC platform only requires engineering a binder to the target protein, as the natural cytokine serves to recruit the degrading receptor. Given these advantages, we were able to express both arms in a single mammalian construct to allow complete assembly of the KineTAC23,24. A His tag was introduced on the knob arm to allow purification of the knob–hole bispecific from contaminants of unwanted hole–hole homodimers that may form. Next, we confirmed that the PD-L1-targeting KineTAC (termed CXCL12–Atz) retains binding to PD-L1 by using biolayer interferometry (BLI; Fig. 1b). Furthermore, an isotype control of the CXCL12 KineTAC, which incorporates a control Fab arm targeting the severe acute respiratory syndrome coronavirus 2 spike protein25, retained binding to endogenous CXCR7 expressed on MDA-MB-231 cells, a breast cancer cell line (Fig. 1c). These data suggest that both anti-PD-L1 and CXCL12 arms of the KineTAC are functional in the bispecific context. To determine the potency of CXCL12–Atz to degrade PD-L1, MDA-MB-231 cells were treated with varying concentrations of the KineTAC. After 24 h of treatment, levels of PD-L1 were quantified by western blotting. Both glycosylated forms of PD-L1 were substantially degraded, with a maximal percent degradation (Dmax) of roughly 70% (Fig. 1d,e). The single-arm Atz Fab or the CXCL12 isotype did not induce the degradation of PD-L1 either alone or in combination, indicating that PD-L1 degradation is dependent on the bispecific KineTAC scaffold (Fig. 1f and Extended Data Fig. 1a,b). Finally, flow cytometry and western blotting of whole-cell lysates verified that PD-L1 is depleted and degraded from the cell surface (Extended Data Fig. 1c,d). Bivalent molecules can induce a ‘hook effect’ at very high concentrations where both target proteins are engaged singly with the bivalent molecule, which could reduce the degradation efficiency26. However, we did not observe reduction of degradation after treatment with high concentrations of CXCL12–Atz (up to 500 nM), which indicates potent degradation over a large dose range (Extended Data Fig. 1e).

We next sought to determine whether the KineTAC degradation platform could be extended to other therapeutically relevant cell surface proteins. First, we targeted human epidermal growth factor receptor 2 (HER2), which is frequently upregulated in cancer and linked to breast cancer invasiveness and tumor progression27. A KineTAC targeting HER2 was produced by incorporating the Fab arm from trastuzumab (Tras; Herceptin), an FDA-approved HER2 inhibitor, into the KineTAC scaffold (termed CXCL12–Tras). Various breast cancer cell lines expressing HER2 were incubated for 24 h with CXCL12–Tras. Substantial degradation of HER2 was observed in MCF7 and MDA-MB-175VII cells, with Dmax values of 51 and 62%, respectively (Fig. 1g,h and Extended Data Fig. 2a). A lower percent degradation was observed in SK-BR-3 cells, likely because HER2 is substantially overexpressed relative to CXCR7 in this cell line28 (Fig. 1h and Extended Data Fig. 2b). Generally, we find that the maximal percent degradation mediated by KineTACs across different cell lines was inversely related to the expression of CXCR7 relative to HER2 (R2 = 0.439; Extended Data Fig. 2c).

Next, we targeted epidermal growth factor receptor (EGFR) for degradation. EGFR is implicated as a driver of cancer progression, and EGFR inhibitors are approved for use in non-small cell lung, colorectal and gastric cancers29,30,31. We developed a KineTAC targeting EGFR by incorporating cetuximab (Ctx; Erbitux), an FDA-approved EGFR inhibitor, into the KineTAC scaffold (termed CXCL12–Ctx). Following 24 h of treatment with CXCL12–Ctx, EGFR levels were dramatically reduced in HeLa cells, with a Dmax of 84% (Fig. 1i,j). This result was recapitulated in various breast and lung cancer cell lines, including MDA-MB-231, A431, NCI-H292, A549 and NCI-H358 cells. As with HER2, the maximum percent degradation was inversely related to the expression of CXCR7 relative to EGFR (R2 = 0.631; Fig. 1j and Extended Data Fig. 2d–h).

We next determined if the KineTACs could be expanded to other tumor-associated membrane proteins . To this end, a KineTAC was constructed from an antibody targeting CUB domain-containing protein 1 (CDCP1)32, which is upregulated in many cancers at levels to over 1 million copies per cell. We observed near complete degradation of CDCP1 after 24 h of treatment in HeLa cells, with a Dmax of 93% (Extended Data Fig. 2i). KineTACs also enabled the degradation of tumor-associated calcium signal transducer 2 (TROP2), the overexpression of which has been linked to tumor progression in a variety of tumors33,34. In MCF7 cells, we observed a Dmax of 51% after treatment with a TROP2-targeting KineTAC (Extended Data Fig. 2j). We then tested whether KineTACs can degrade the checkpoint protein PD-1 in CD8+ T cells isolated from primary human peripheral blood mononuclear cells (PBMCs). T cells were first activated to induce upregulation of PD-1 on the cell surface along with other activation markers (Extended Data Fig. 3a,b). The activated T cells were then treated for 24 h with a PD-1-targeting KineTAC, which incorporated the antibody sequence for nivolumab (Nivo; Opdivo), an FDA-approved PD-1 inhibitor (termed CXCL12–Nivo). Following treatment with CXCL12–Nivo, the proportion of PD-1 on the surface of T cells was reduced by 82% compared to treatment with the Nivo isotype control, which is known to induce slight internalization of PD-1 (Fig. 1k and Extended Data Fig. 3c)35. Overall, these results on six different membrane proteins demonstrate the generality of the KineTAC platform for degrading a variety of cell surface proteins (Supplementary Table 1).

Next, we evaluated how properties, such as receptor signaling, binding affinity and binding epitope, of the KineTAC affected degradation. In addition to binding the recycling receptor CXCR7, CXCL12 binds the signaling receptor CXCR4, which causes downstream signaling followed by receptor internalization and degradation. It is possible that signaling through CXCR4 could be counterproductive when using KineTACs to target cancer drivers because CXCR4 overexpression and agonism are linked to tumor metastasis36. To test if CXCR4 signaling is critical for the KineTAC mechanism of action, we produced known antagonistic variants of CXCL12 (∆KP, ∆KPVS and R8E) that retain high binding affinity to both CXCR7 and CXCR4 but prevent CXCR4 signaling19,37. These were incorporated into the KineTAC scaffold with Atz. Following 24-h treatment in MDA-MB-231 cells, all three antagonistic variants retained the ability to degrade PD-L1 to a similar degree as CXCL12 wild type (CXCL12WT; Fig. 2a,b). These data suggest that CXCL12 signaling through CXCR4 is not critical for degradation.

Fig. 2: Requirements for efficient KineTAC-mediated degradation of target proteins.
figure 2

a,b, Treatment with 100 nM antagonistic (∆KP, ∆KPVS and R8E) CXCL12–Atz variants for 24 h in MDA-MB-231 cells shows no significant difference in PD-L1 levels compared to agonistic CXCL12WT–Atz at 50 nM (P = 0.1432, 0.1222 and 0.5016, respectively). At 100 nM, CXCL12R8E–Atz showed significantly greater PD-L1 degradation than CXCL12WT–Atz (P = 0.0162), while CXCL12∆KP–Atz and CXCL12∆KPVS–Atz were unchanged (P = 0.0609 and 0.7538, respectively); n = 3 biologically independent experiments for CXCL12WT–Atz, and n = 2 for CXCL12–Atz variants. c, PD-L1 levels in MDA-MB-231 cells after treatment with CXCL12–Atz wild type or alanine mutants (100 nM) for 24 h. df, Correlation of PD-L1 levels as calculated by densitometry and Kd (d), kon (e) or koff (f). Wild-type Atz is indicated in red; n = 3 biologically independent experiments. g, PD-L1 levels after treatment with 100 nM CXCL12–Atz or pH-sensitive binder CXCL12–BMS936559 for 24 h in MDA-MB-231 cells. h, HER2 levels after treatment with 100 nM CXCL12–Tras or CXCL12–Ptz in MCF7 cells demonstrate that different epitope binders affect the half-maximum degradation concentration (DC50) of HER2 degradation, while the Dmax at 100 nM was unchanged (P = 0.7758); n = 2 biologically independent experiments. i, EGFR levels after treatment with 100 nM CXCL12–Ctx, CXCL12–Depa, CXCL12–Nimo, CXCL12–Matu, CXCL12–Neci or CXCL12–Pani in HeLa cells demonstrate that there is dependence on EGFR binding epitope for degradation efficiency. CXCL12–Ctx, CXCL12–Pani, CXCL12–Neci and CXCL12–Matu significantly degrade EGFR compared to isotype controls (P = 0.0004, 0.016, 0.006 and 0.009, respectively), while CXCL12–Depa and CXCL12–Nimo do not cause significant EGFR degradation (P = 0.7619 and 0.3573, respectively); n = 3 biologically independent experiments. Densitometry was used to calculate protein levels, and data were normalized to PBS control. Data are represented as mean values, and error bars represent the standard deviation of biological replicates. P values were determined by unpaired two-tailed t-tests. Linear regression analysis using GraphPad Prism was used to calculate the coefficient of determination (R2) to determine the strength of linear correlation.

Source data

To test how the affinity for target protein PD-L1 affects degradation efficiency, we systematically produced disruptive alanine mutations in the complementary-determining regions of Atz known to contact PD-L1 (ref. 38). These mutant Atz variants bound with a wide range of binding affinities (KD values from 0.33 to 458 nM) and corresponding kinetic parameters (kon and koff), as measured with BLI (Supplementary Table 2). The Atz mutants were then introduced into the KineTAC scaffold with CXCL12WT and tested for their ability to degrade PD-L1 (Fig. 2c and Extended Data Fig. 4a). We found that degradation inversely correlates with the Kd (R2 = 0.638) and even more strongly to dissociation rate (koff, R2 = 0.804) but not to the association rate (kon, R2 = 0.036; Fig. 2d–f). The wild-type Atz had the highest binding affinity and induced the greatest level of PD-L1 degradation. To test the importance of CXCL12–CXCR7 binding affinity, we produced known N-terminal antagonistic variants of CXCL12 with a range of binding affinities to CXCR7 (Kd or half-maximum inhibitory concentration (IC50) values ranging from 0.014 to 28 nM). These were incorporated into the KineTAC scaffold with Atz and tested for their ability to degrade PD-L1 (Extended Data Fig. 4b)39. Remarkably, the degradation potency showed only a weak trend to the IC50 against CXCR7 (R2 = 0.206; Extended Data Fig. 4c). Thus, the levels of degradation appear more dependent on binding affinity of the antibody arm to the target protein than on binding affinity of the CXCL12 variant to CXCR7. Because the effects of these mutations are almost entirely on koff, this suggests that higher residence time on the target is more important than higher residence time on CXCR7 for robust degradation.

To determine whether a pH-dependent antibody binder against the target protein would affect degradation, BMS936559, a PD-L1 antibody reported to release PD-L1 under acidic conditions (pH < 6.0)40, was incorporated into the KineTAC scaffold. The maximal degradation efficiency with CXCL12–BMS936559 was slightly reduced compared to CXCL12–Atz, suggesting that unlike sweeping antibodies, the CXCL12-based KineTAC is not advantaged by low-pH release (Fig. 2g). This result is not due to differences in Kd because Atz and BMS936559 are reported to have similar binding affinities to PD-L1 (ref. 41).

To investigate how the binding epitope on the protein of interest impacts degradation, we generated KineTACs with different HER2- and EGFR-targeting antibodies. For HER2, pertuzumab (Ptz; Perjeta), which binds a distinct epitope from Tras on HER2 (ref. 42), was introduced into the KineTAC scaffold (termed CXCL12–Ptz). Following 24-h treatment of MCF7 cells, we found that the IC50 for CXCL12–Tras is significantly lower than that for CXCL12–Ptz while reaching the same Dmax, indicating that the binding epitope can alter the dose response to KineTACs (Fig. 2h). This effect is not due to differences in binding affinity, as Tras and Ptz have virtually identical Kd values of 1.43 and 1.92 nM, respectively43. For EGFR, we introduced five additional anti-EGFR binders (depatuxizumab (Depa), nimotuzumab (Nimo), panitumumab (Pani; Vectibix), necitumumab (Neci; Portrazza) and matuzumab (Matu))44 into the KineTAC scaffold that bind different epitopes on EGFR domains 2 and 3 but have similar affinities (Extended Data Fig. 4d). Following 24-h treatment of HeLa cells, we observed that KineTACs containing Neci, Matu and Pani caused significant degradation of EGFR similar to CXCL12–Ctx, while those containing Depa and Nimo did not substantially degrade EGFR (Fig. 2i). The degradation efficiency of these antibodies was not correlated to their respective binding affinities (R2 = 0.008; Extended Data Fig. 4e). It is also known that Depa and Nimo bind further from the cell surface than the binders that produced more effective KineTACs. These data highlight the dependence on the binding epitope on the target for KineTAC-mediated degradation.

Glycosylation at N297 in the Fc region can impart greater circulating half-life into IgG1 (ref. 45). To test the impact of glycosylation on KineTACs, we introduced N297 into the CXCL12–Atz scaffold and tested the degradation efficiency between the glycosylated and aglycosylated forms. We found that glycosylation at N297 does not substantially impact PD-L1 degradation levels (Extended Data Fig. 4f). Thus, it is possible to incorporate N297 glycosylation for improved pharmacokinetic properties of IgG-containing KineTACs without impacting degradation potency.

Lastly, we determined the sensitivity to orientation of the two binding arms in the KineTAC construct by comparing the parallel Fc-based bispecific to head–tail-linked tandem binders. The latter was produced by fusing CXCL12 to the N terminus of the Atz Fab heavy chain via a flexible avidin tag linker and coexpressing with Atz Fab light chain (Extended Data Fig. 4g). The CXCL12–Atz Fab fusion retained high binding to PD-L1, as measured by BLI (Extended Data Fig. 4h). After 24 h of treatment in MDA-MB-231 cells, the levels of PD-L1 degradation for the tandem Fab fusion construct were substantially less (Dmax = 20%; Extended Data Fig. 4i) than for the Fc-based bispecific (Dmax = 70%; Fig. 1d,e). We also tested an IgG fusion construct where CXCL12 was N-terminally fused to either the heavy chain or light chain of the Fab arm (Extended Data Fig. 4j). The bispecific knob-in-hole IgG construct was significantly better than the bivalent IgG fusions (Extended Data Fig. 4k). Our data suggest that the parallel orientations were more effective than the tandem orientations. Overall, our data highlight that construct geometry is an important variable for KineTAC degradation efficiency.


We next sought to evaluate the mechanism of KineTAC-mediated degradation. To determine whether KineTACs catalyze degradation via the lysosome or proteasome, MDA-MB-231 cells were pretreated with either medium alone, bafilomycin (an inhibitor of lysosome acidification) or MG132 (a proteasome inhibitor). After 1 h of drug treatment, cells were treated with CXCL12–Atz for 24 h. Bafilomycin pretreatment inhibited degradation of PD-L1, while MG132 had no effect, suggesting that KineTACs mediate degradation via delivery of target proteins to the lysosome (Fig. 3a). Immunofluorescence microscopy revealed virtually complete removal of EGFR from the cell surface following 24 h of CXCL12–Ctx treatment compared to treatment with PBS control, further highlighting that KineTACs induce robust internalization of target proteins (Fig. 3b). KineTAC-mediated degradation occurs in a time-dependent fashion, beginning 6 h after treatment with CXCL12–Atz, with the levels of PD-L1 continuing to decrease over time to near complete degradation by 48 h (Fig. 3c,d).

Fig. 3: KineTACs mediate target degradation in a highly selective, lysosomal-, time- and CXCR7-dependent manner.
figure 3

a, Pretreatment for 1 h with either 100 nM bafilomycin or 5 µM MG132 in MDA-MB-231 cells followed by 24 h of treatment with 100 nM CXCL12–Atz indicates that CXCL12–Atz degrades PD-L1 in a lysosome-dependent manner. b, Confocal microscopy images of HeLa cells treated for 24 h with 100 nM CXCL12–Ctx show near complete removal of EGFR. c,d, Time-course experiment shows increased PD-L1 degradation over time after treatment with 100 nM CXCL12–Atz. e, siRNA knockdown of CXCR4 in HeLa cells after a 48-h transfection. f, EGFR levels are unchanged after siRNA knockdown of CXCR4 followed by 24 h of treatment with 100 nM CXCL12–Ctx in HeLa cells. g, PD-L1 in MDA-MB-231 cells or EGFR in HeLa cells is significantly degraded after 24 h of treatment with 100 nM CXCL11–Atz (P = 0.0002) or CXCL11–Ctx (P = 0.0014) compared to isotype controls. No significant difference was observed between CXCL11–Atz and CXCL12–Atz (P = 0.2643), while CXCL12–Ctx caused greater EGFR degradation than CXCL11–Ctx (P = 0.0146); n = 3 biologically independent experiments. h,i, Fold change in abundance of surface-enriched (h) or whole-cell (i) MDA-MB-231 proteins detected using quantitative proteomics analysis after 48 h of treatment with 100 nM CXCL12–Atz reveals highly selective PD-L1 degradation. j,k, Fold change in abundance of surface-enriched (j) or whole-cell (k) HeLa proteins detected using quantitative proteomics analysis after 48 h of treatment with 100 nM CXCL12–Ctx reveals highly selective EGFR degradation. Data are the mean of n = 2 biological independent experiments and two technical replicates. Proteins showing a greater than twofold change from PBS control with a significance of P < 0.01 were considered significantly changed. l, In vitro potency of CXCL12–Tras in MDA-MB-175VII cells demonstrates superior cell killing versus Tras IgG; n = 2 biologically independent experiments and three technical replicates. m, Quantification of CXCL12–Tras plasma levels in male nude mice injected intravenously at 5, 10 or 15 mg per kg (body weight); n = 3 different mice per group. Densitometry was used to calculate protein levels, and data were normalized to whole protein levels. Data are represented as mean values, and error bars represent the standard deviation of biological replicates. P values for CXCL11 data were determined by one-way ANOVA with Sidak’s multiple comparisons test. P values for cell viability assays were determined by unpaired two-tailed t-tests at each indicated dose.

Source data

We next wished to determine which of the two receptors for CXCL12 is more important for degrading surface proteins in the context of KineTACs. We directly probed the role of CXCR4 by using RNA interference to knockdown the levels of CXCR4 in HeLa cells. After a 48-h transfection with short interfering RNA (siRNA) pools targeting CXCR4 or a nonspecific control, cells were treated with CXCL12–Ctx for 24 h (Fig. 3e). Western blotting analysis revealed that EGFR degradation levels were unchanged with CXCR4 knockdown. These data indicate that CXCR4 is not necessary for efficient degradation and suggest that KineTACs operate through CXCR7-mediated internalization (Fig. 3f). Furthermore, KineTACs bearing CXCL11, a chemokine that specifically binds CXCR7 and CXCR3 but not CXCR4 (ref. 46), are capable of efficiently degrading both PD-L1 and EGFR (Fig. 3g and Extended Data Fig. 5a–d). A KineTAC containing vMIPII, a viral chemokine that targets CXCR7 along with other chemokine receptors, also efficiently degraded PD-L1 (Extended Data Fig. 5e). Our results suggest that CXCR7 is the prime receptor responsible for KineTAC-mediated degradation and demonstrate the exciting opportunity for using alternative cytokines, such as CXCL11 and vMIPII, in the KineTAC scaffold to degrade target proteins.

Next, we used quantitative mass spectrometry to determine changes across the proteome that occur following KineTAC treatment. Both the surface-enriched and whole-cell lysates were analyzed following 48 h of treatment with the PD-L1-targeting KineTAC CXCL12–Atz or PBS (control) in MDA-MB-231 cells. Remarkably, PD-L1 was virtually the only protein downregulated after CXCL12–Atz treatment compared to after control treatment for the surface-enriched sample (Fig. 3h). Whole-cell proteomics also revealed that no major changes occurred (Fig. 3i and Supplementary Table 3). PD-L1 was not detected in the whole-cell sample, likely due to low abundance of cell surface proteins relative to cytosolic proteins.

Similar results were observed for HeLa cells treated with the EGFR-targeting KineTAC CXCL12–Ctx compared to cells treated with PBS control (Fig. 3j,k). There were no major changes across the proteome, except for a dramatic reduction in EGFR in both the surface-enriched and whole-cell lysates (Fig. 3j,k and Supplementary Table 4). Moreover, CXCR4 and CXCR7 peptide ID numbers were not altered in the surface-enriched sample, and CXCR4 ID numbers were also unchanged in the whole-cell sample, indicating that treatment with KineTAC does not significantly impact CXCR4 or CXCR7 levels. CXCR7 peptide ID numbers were not identified in the whole-cell sample, likely due to low abundance relative to cytosolic proteins. Furthermore, protein levels of GRB2 and SHC1, which are known interacting partners of EGFR47,48, were also not significantly changed. However, we did observe the significant downregulation of ABCD1 and BGH3 in the PD-L1 and EGFR datasets, respectively. These appear to be unrelated to either target protein or CXCR7, suggesting that their reduction was independent of the KineTAC mechanism of action. Together, these data suggest that CXCL12-based KineTACs are highly selective for the target protein without substantially altering levels of the degrading receptor or known interacting partners to EGFR. Interestingly, a previously published proteomics dataset of LYTAC-mediated EGFR degradation identified 25 proteins significantly up- or downregulated following LYTAC treatment2. In our HeLa cell dataset, we observed that 89% of the proteins identified in the total LYTAC dataset, including 23 of the 25 proteins seen to significantly change after LYTAC treatment; these 23 proteins were unchanged in the KineTAC dataset (Extended Data Fig. 6a,b). These data suggest that the CXCL12-based KineTAC exhibits more selective EGFR degradation.

Next, we sought to elucidate whether KineTAC-mediated degradation could impart functional cellular consequences. Tras inhibits HER2 function, and MDA-MB-175VII breast cancer cells, which overexpress HER2, will undergo apoptosis after 5 d of treatment with Tras49. We compared the abilities of CXCL12–Tras and Tras IgG to kill MDA-MB-175VII cells by using a modified MTT assay. We found that the CXCL12–Tras KineTAC induced significantly greater cell death than Tras IgG, Tras Fab or CXCL12 isotype alone (Fig. 3l and Extended Data Fig. 6c,d). Additionally, non-small cell lung cancer NCI-H358 cells, which overexpress EGFR, undergo significantly greater apoptosis when treated with the CXCL12–Ctx KineTAC than when treated with Ctz IgG (Extended Data Fig. 6e). These data demonstrate that KineTAC-mediated degradation can provide functional advantages over traditional antibody therapeutics, which bind and inhibit but do not degrade.

We next asked whether KineTACs would have similar antibody clearance to IgGs in vivo. To this end, male nude mice were injected intravenously with 5, 10 or 15 mg per kg (body weight) CXCL12–Tras (N297 glycosylation present), which is a typical dose range for antibody xenograft studies. Western blotting analysis of plasma antibody levels revealed that the KineTAC remained in plasma up to 10 d after injection with a half-life of 8.7 d, which is comparable to the reported half-life of IgGs in mice (Fig. 3m and Extended Data Fig. 7a)50. Given the high homology between human and mouse CXCL12, we tested whether human CXCL12 isotype was cross-reactive. In fact, human CXCL12 isotype bound to mouse cell lines MC38 and CT26 that endogenously express mouse CXCR7 (Extended Data Fig. 7b,c). Together, these results demonstrate that KineTACs have favorable pharmacokinetics and are not being rapidly cleared despite cross-reactivity with mouse CXCR7 receptors. Because Atz is also known to be cross-reactive, the ability of CXCL12–Atz to degrade mouse PD-L1 was tested in both MC38 and CT26 cells. Indeed, CXCL12–Atz mediated near complete degradation of mouse PD-L1 in both cell lines (Extended Data Fig. 7d–f). Thus, we observe a long in vivo half-life of KineTACs despite cross-reactivity and binding to mouse CXCR7 receptors.

Having demonstrated the ability of KineTACs to mediate cell surface protein degradation, we next asked whether KineTACs could also be applied toward the degradation of soluble extracellular proteins. Antibodies binding to soluble ligands, such as vascular endothelial growth factor (VEGF) and tumor necrosis factor-α (TNF-α), are of tremendous therapeutic importance51. Thus, we investigated whether KineTACs could promote cellular uptake of either VEGF or TNF-α (Fig. 4a). VEGF was targeted by incorporating bevacizumab (Beva; Avastin), an FDA-approved VEGF inhibitor, into the KineTAC scaffold (termed CXCL12–Beva). HeLa cells were incubated with Alexa Fluor 647-labeled VEGF (VEGF–647) alone or VEGF–647 plus CXCL12–Beva for 24 h. Following treatment, flow cytometry analysis showed a robust 32-fold increase in cellular fluorescence when VEGF–647 was co-incubated with CXCL12–Beva but not when co-incubated with Beva isotype IgG (Fig. 4b,c). To ensure that the increased cellular fluorescence was due to intracellular uptake of VEGF–647 and not surface binding, we determined the effect of a trypsin lift after treatment, which degrades any cell surface-bound VEGF–647 (Extended Data Fig. 8a,b). We found that there was no significant difference in cellular fluorescence levels between trypsin- and normal-lifted cells treated for 24 h with CXCL12–Beva (Fig. 4d). We did observe a difference between Beva isotype-treated cells, which may reflect slight nonspecific internalization that occurs with the trypsin lift. However, this difference was not observed for CXCL12–Beva. Pretreatment with bafilomycin, a lysosome inhibitor, also impaired the ability of CXCL12–Beva to take up extracellular VEGF–647, as observed by a decrease in cellular fluorescence (Fig. 4e). VEGF–647 uptake was not completely impaired, likely due to bafilomycin blocking trafficking from the endosome to the lysosome, which would still allow some VEGF–647 to be endocytosed but not degraded. Together, these data suggest that KineTACs successfully mediate the intracellular uptake of extracellular VEGF and delivery to the lysosome. KineTAC-mediated uptake of VEGF occurs in a time-dependent manner, similar to membrane protein degradation, with robust internalization beginning at 6 h and reaching steady state by 24 h (Fig. 4f). Furthermore, the levels of VEGF uptake are dependent on the KineTAC:ligand ratio and saturate at ratios greater than 1:1 (Fig. 4g). We next tested the ability of CXCL12–Beva to promote uptake in other cell lines, including breast and lung cancer lines, and found that these cells also significantly take up VEGF (Fig. 4h). Moreover, the extent of uptake was positively correlated with ACKR3 transcript levels in these cells (R2 = 0.555; Extended Data Fig. 8c). These data suggest that KineTACs directed against soluble ligands can promote uptake of extracellular soluble ligands and that cell lines expressing higher levels of CXCR7 lead to greater uptake.

Fig. 4: KineTACs enable intracellular uptake of soluble extracellular proteins.
figure 4

a, Schematic of KineTAC concept for targeting extracellular proteins for lysosomal degradation. The red ball represents an extracellular protein bound to the KineTAC bi-specific. b, Representative flow cytometry results showing a shift in MFI in HeLa cells treated for 24 h with 50 nM CXCL12–Beva and 25 nM VEGF–647 compared to treatment with VEGF alone. c, Bar graph of flow cytometry results demonstrating significant uptake of VEGF–647 in HeLa cells following 24 h of treatment with 50 nM CXCL12–Beva compared to treatment with isotype control (P < 0.0001). Beva isotype control showed no difference compared to treatment with VEGF–647 alone (P = 8421); n = 3 biologically independent experiments. d, Comparison of HeLa cells lifted with versene (normal lift) or 0.25% trypsin-EDTA (trypsin lift) following 24 h of treatment with 50 nM CXCL12–Beva or isotype controls and 25 nM VEGF–647. No significant change in MFI for CXCL12–Beva treatment (P = 0.1121) suggests that fluorescence shift represents accumulation of intracellular VEGF–647 and not simple surface binding; n = 3 biologically independent experiments. e, Summary of flow cytometry results demonstrating a significant decrease in MFI in HeLa cells following pretreatment with 100 nM bafilomycin and 24 h of treatment with 50 nM CXCL12–Beva and 25 nM VEGF–647 (P = 0.0005) compared to no pretreatment with bafilomycin; n = 3 biologically independent experiments. f, Time-course experiment showing increase in VEGF–647 uptake over time in HeLa cells treated with 50 nM CXCL12–Beva and 25 nM VEGF–647; n = 3 biologically independent experiments. g, HeLa cells treated for 24 h with varying ratios of CXCL12–Beva to VEGF at constant 25 nM VEGF–647 demonstrate that increasing the KineTAC:VEGF ratio increases VEGF uptake; n = 3 biologically independent experiments. h, Panel of cell lines for VEGF–647 uptake experiments demonstrating significant VEGF–647 uptake following CXCL12–Beva treatment compared to following treatment with Beva isotype or VEGF alone in MDA-MB-231 (P = 0.048), NCI-H292 (P = 0.049), MCF7 (P = 0.017) and HeLa (P = 0.0005) cells; n = 3 biologically independent experiments. MFI was measured using live-cell flow cytometry. Data are represented as mean values, and error bars represent the standard deviation of biological replicates. P values were determined by unpaired two-tailed t-tests or one-way ANOVA with Sidak’s multiple comparisons test. Fold changes are reported relative to incubation with soluble ligand alone.

Source data

We next targeted TNF-α by incorporating adalimumab (Ada; Humira), an FDA-approved TNF-α inhibitor, into the KineTAC scaffold (termed CXCL12–Ada). Following 24 h of treatment in HeLa cells, a significant 8.5-fold increase in cellular fluorescence was observed when TNF-α–647 was co-incubated with CXCL12–Ada compared to when TNF-α–647 was co-incubated with Ada isotype (Extended Data Fig. 9a,b). Consistent with the VEGF uptake experiments, TNF-α uptake was dependent on the KineTAC:ligand ratio (Extended Data Fig. 9c). Thus, KineTACs can robustly mediate the intracellular uptake of soluble ligands, substantially expanding the target scope of KineTAC-mediated targeted degradation.

Finally, we asked whether alternative cytokine receptors could be co-opted by KineTACs to mediate the clearance of target proteins. We hypothesized that KineTACs bearing interleukin-2 (IL-2) cytokine could engage the IL-2 receptor complex (IL-2R), which comprises CD25, IL-2RB and IL-2RG, as a degrading receptor. CD25 is known to internalize and recycle back to the cell surface after heterotrimer formation, while IL-2RB and IL-2RG are internalized and degraded via the lysosome after IL-2 binding (Extended Data Fig. 10a)52. We generated knob-in-hole KineTACs in which human IL-2 cytokine was N-terminally fused to the knob IgG1 Fc domain, and the hole Fc arm contained the Fab antibody sequence for Nivo, which binds PD-1. Activated T cells were then treated for 24 h with the IL-2-bearing PD-1 KineTAC (termed IL-2–Nivo). Following treatment with IL-2–Nivo, cell surface PD-1 levels were significantly reduced, with a Dmax of 86.7%, compared to those observed after treatment with Nivo isotype control (Extended Data Fig. 10b). These data suggest that other cytokines can be co-opted with KineTACs, expanding the repertoire for cell-selective targeted protein degradation. We note that bispecific constructs currently called immunocytokines link a cytokine to a targeting antibody with the intent to deliver cytokines such as IL-2 to T cells in the tumor microenvironment. In addition to selective cytokine delivery, our data suggest that these may also inadvertently degrade the target protein if it is also present on the T cells, which could be an advantage in targeting PD-1 with an IL-2-bearing KineTAC.


Click to comment

Leave a Reply

Your email address will not be published.



Copyright © 2022 Theme by The Nitesh Arya.