A genome-integrated platform for conditional protein depletion

To establish the AIDE system in Chlamydia trachomatis (Ctr), we designed a strategy to integrate degron sequences into chlamydial target effector proteins (Ctr-AIDE). Central to this system is a 7-kDa mAID degron, fused to target effectors and recognized by the mutant host F-box receptor OsTIR1(F74G), which forms the SCF E3 ligase complex and thus directs K48-linked ubiquitination upon addition of the synthetic auxin 5-Ph-IAA. To enable precise degron integration into Ctr, we combined AID2 with the FRAEM (Fluorescence-Reported Allelic Exchange Mutagenesis) system, a homologous recombination-based method for seamless genome editing⁴¹. This integration yielded Ctr-AIDE, a platform for rapid and reversible depletion of secreted effectors (Fig. 1A).

The alternative text for this image may have been generated using AI.

A Schematic of Ctr-AIDE strategy, combining the FRAEM genome-editing platform (left) with the AID2 TPD system (right). Cdu1 is shown as proof of concept. Created in BioRender. Zhang, H. (2026) https://BioRender.com/t5v0ec7. B Top: Plasmid constructs for FRAEM-mediated homologous recombination in Ctr serovar L2. The targeted genomic region is marked by lines. pL2ori represents the putative chlamydial plasmid ori, and pBR322 (Ori) is included to facilitate cloning in E. coli. pCdu1, native Cdu1 promoter; Nmp, Neisseria meningitidis promoter; aadA, spectinomycin resistance gene; Bottom: Ctr-AIDE degradation/re-expression strategy. Bars indicate treatment with DMSO (control) or 5-Ph-IAA (1 µM). All samples were harvested at 28 hours post-infection (hpi) for downstream analysis. C Ctr-AIDE-mediated regulation of Cdu1-mAID. Cells were infected with indicated strains at MOI = 1. Cdu1 (anti-FLAG) and OsTIR1 mutants (TIR1-F74G denotes OsTIR1(F74G)-9×Myc, and TIR1-7K/10 K/74 G denotes OsTIR1(E7K/E10K/F74G)-9×Myc; anti-Myc) levels were monitored, with Chlamydial Major Outer Membrane Protein (OmpA) as a loading control and GAPDH as a host cell control. Asterisks denote putative truncated or fragmented forms of Cdu1. D Quantification of Cdu1 degradation. FLAG signal (normalized to OmpA, mean ± SD) from 3 independent immunoblots (one shown in C). Statistical significance determined by two-tailed paired t‑test (***p p p p values were provided in Supplementary Data 3). OsTIR1_Ina, -mAID degron, -5-Ph-IAA: controls with dysfunctional E3 ligase OsTIR1(E7K/E10K/F74G), lacking mAID tagging, and lacking 5-Ph-IAA treatment, respectively. All experiments were performed ≥3 times with consistent results. Source data are provided as a Source Data file.

For several reasons, we selected the chlamydial deubiquitinase Cdu1 to establish the Ctr-AIDE strategy: (i) Cdu1 is a highly active K48 deubiquitinase that upon AIDE (that ubiquitinated K48 to induce degradation) could lower the efficiency of degradation; (ii) The auto-acetylation of Cdu1 at lysine residues has been implicated in ubiquitination resistance of the protein³⁸; (iii) As a membrane integrated protein, degradation of Cdu1 is expected to be particularly challenging; (iv) Rapid Cdu1 protein degradation and re-expression could be a way to shed light on the effect of the deubiquitinating and acetyltransferase activities of Cdu1 in ongoing infections. To generate the Cdu1 degradation construct, a knock-in cassette was designed in the suicide plasmid pKW-L2⁴¹, consisting of (i) 3 kb homology sequences flanking the cdu1 locus; (ii) an mAID-FLAG sequence fused to the C-terminus of Cdu1 via a flexible linker (2x GGGS), ensuring cytosolic accessibility given that the Cdu1 C-terminus faces the host cytosol³⁶; and (iii) a GFP selection cassette with spectinomycin resistance gene (aadA-GFP, Fig. 1B). This design enables allelic replacement of endogenous cdu1 with the mAID-FLAG-tagged variant, ensuring native expression regulation and eliminating reliance on plasmid-based expression. Ctr recombinants were initially screened for loss of the plasmid pKW-L2 (RFP^–) and the expression of GFP indicative of successful recombination. In addition, mAID tagging was confirmed by immunoblotting (Supplementary Fig. 1A). A matched Cdu1-FLAG control strain that lacks the mAID degron was constructed in parallel for comparative analysis (Supplementary Fig. 1A).

We established stable HeLa cell lines expressing either functional OsTIR1(F74G) or catalytically impaired OsTIR1(E7K/E10K/F74G)⁴², and validated AID2 functionality by mAID-GFP degradation. The impaired OsTIR1(E7K/E10K/F74G) mutant served as a reliable degradation-negative control (Supplementary Fig. 2). Using this system, we investigated Ctr-AIDE-mediated depletion of the effector Cdu1 (Fig. 1B). No K48 ubiquitin was detected on the inclusions in the Ctr-AIDE strain indicating that Cdu1-mAID functions in protecting the inclusion from ubiquitination. Strong K48 ubiquitination was visible on the inclusion already 5 min after addition of the 5-Ph-IAA inducer (Supplementary Fig. 3) suggesting that OsTIR1(F74G) is recruited to Cdu1-mAID and induces its ubiquitination. Surprisingly, strong degradation of Cdu1-mAID was visible already 15 minutes after addition of the 5-Ph-IAA inducer and degradation was complete within 30 min (Fig. 1C, D). Degradation depended on the mAID tag since Cdu1 remained stable in (i) cells expressing the catalytically impaired OsTIR1(E7K/E10K/F74G) variant⁴², (ii) strains expressing non-degron Cdu1-FLAG constructs, and (iii) untreated conditions. Importantly, degradation was fully reversible, with Cdu1-mAID re-accumulating partially within 15 min and completely within 2 h following 5-Ph-IAA washout after 2 h of treatment (Fig. 1C, D). Immunofluorescence microscopy confirmed rapid loss of inclusion membrane-localized Cdu1 after 5-Ph-IAA treatment and its restoration upon 5-Ph-IAA washout (Fig. 2A). Spatial resolution was further validated by 4 × expansion microscopy⁴³, which confirmed specific depletion of inclusion membrane-associated Cdu1 while pools in the chlamydial cytoplasm remained unaffected, highlighting the system’s spatial precision (Fig. 2B). To assess target specificity of Ctr-AIDE for mAID-tagged effector, we monitored the untagged inclusion membrane protein IncA during rapid Cdu1 depletion. Immunoblotting showed that IncA expression remained unchanged while Cdu1-mAID was rapidly degraded (Supplementary Fig. 4), indicating that Ctr-AIDE selectively degrades mAID-tagged proteins.

The alternative text for this image may have been generated using AI.

A Immunofluorescence microscopy confirms Cdu1 depletion dynamics. Cdu1 (green, anti-FLAG), chlamydial inclusion (Hsp60, red), and DNA (DAPI, blue). The white square in the left panel marks the region enlarged in the right panel. Scale bar = 10 µm. B Expansion microscopy (4× expansion) resolving Cdu1 localization patterns under undegraded and degraded conditions. Scale bar = 10 µm. C Cdu1 degradation depends on ubiquitin-proteasome and p97 pathways. Host cells were pretreated with inhibitors for 3 h (including 1 hour during 5-Ph-IAA treatment) with pan-E1 (TAK-243, 1 µM), proteasome (MG-132, 10 µM; bortezomib, 1 µM), or p97 (CB-5083, 10 µM) inhibitors. Degradation was blocked despite 1-hour 5-Ph-IAA exposure. D Quantification of inhibitor effects on Cdu1 degradation. FLAG levels (normalized to OmpA, mean ± SD) from three independent immunoblot replicates (one representative in C). Significance assessed by one-way ANOVA with Tukey’s Multiple comparisons test. (****p p values were provided in Supplementary Data 3). E Immunofluorescence validation of inhibitors. Inclusion-localized Cdu1 signal persists in treated cultures. Scale bar = 10 µm. All experiments were replicated ≥3 times with consistent results. Source data are provided as a Source Data file.

Mechanistic dissection confirmed that Cdu1-mAID degradation depends on the host ubiquitin-proteasome pathways: degradation was blocked by the inhibition of the central Ubiquitin-activating enzyme 1 (TAK-243), and the proteasomal activity (MG-132, bortezomib). Interestingly, inhibition of the p97 ATPase with CB-5083 blocks Cdu1-mAID degradation. p97 functions by extracting misfolded or damaged proteins from membranes, a process triggered by ubiquitination⁴⁴, suggesting that degradation of inclusion membrane proteins involves p97, potentially through extraction of membrane proteins from the inclusion prior to proteasomal degradation. (Fig. 2C–E).

To assess the functionality of Ctr-AIDE in different cell lines, Cdu1 degradation was tested in A-375 (melanoma), HCT 116 (colon carcinoma), and U-2 OS (osteosarcoma). All lines exhibited degradation kinetics comparable to HeLa (cervical carcinoma) cells (Supplementary Fig. 5, 6), with HeLa showing the highest proteasomal degradation efficiency (Supplementary Fig. 5B). These results confirm that Ctr-AIDE functions robustly across diverse host cell types, with degradation fidelity dependent on the host ubiquitin-proteasome machinery rather than cell type-specific factors.

These results validate Ctr-AIDE as a genome-integrated AID2 platform that enables precise, conditional depletion of Chlamydia effectors across diverse host cell lines while preserving native expression dynamics. By eliminating plasmid dependencies and leveraging the host ubiquitin-proteasome pathways, Ctr-AIDE establishes a framework for dissecting essential bacterial virulence factors during infection.

Ctr-AIDE functions in primary cells

To validate the broad applicability of Ctr-AIDE, we established the Ctr-AIDE system in primary cells. Primary Murine Reproductive Tract (PMRT) organoids, derived from whole female reproductive tract digests, were transduced with lentivirus encoding OsTIR1(F74G) or OsTIR1(E7K/E10K/F74G) and selected for stable expression (Supplementary Fig. 7). Cells derived from these organoids were then grown as 2D monolayers to facilitate 5-Ph-IAA treatment when used for Ctr-AIDE degradation assays. In PMRT cells expressing OsTIR1(F74G), we observed rapid, 5-Ph-IAA-dependent depletion of mAID-tagged Cdu1 comparable to cancer cell lines: immunoblotting confirmed near-complete Cdu1 loss within 1 h of 5-Ph-IAA treatment, while controls (Primary cells expressing dysfunctional OsTIR1(E7K/E10K/F74G), DMSO treated cells, and non-degron Cdu1 constructs) showed no degradation (Fig. 3A, B). Immunofluorescence microscopy further demonstrated auxin-driven disappearance of inclusion membrane-localized Cdu1-mAID and its restoration within 2 h after 5-Ph-IAA washout (Fig. 3C). These results confirm that Ctr-AIDE operates robustly across both cancer and primary cell models, further prove its versatility for studying Chlamydia effector dynamics in physiologically relevant environments.

The alternative text for this image may have been generated using AI.

A Immunoblot of the Ctr-AIDE-mediated regulation of Cdu1 levels in PMRT cells. Degradation treatment as in Fig. 2, but a higher infection load (MOI = 10). Cdu1 and OsTIR1 mutants (TIR1-F74G denotes OsTIR1(F74G)-9×Myc, and TIR1-7K/10 K/74 G denotes OsTIR1(E7K/E10K/F74G)-9×Myc) were detected with anti-FLAG and anti-Myc antibodies; OmpA and GAPDH served as loading controls. Asterisks denote putative truncated or fragmented forms of Cdu1. B Quantification of Cdu1 degradation in primary cells. OsTIR1_Ina, -mAID degron, -5-Ph-IAA: controls with dysfunctional E3 ligase OsTIR1(E7K/E10K/F74G), lacking mAID tagging, and lacking 5-Ph-IAA treatment, respectively. FLAG levels (normalized to OmpA, mean ± SD) from three independent replicates (one shown in A). Significance assessed by two-tailed paired t‑test (*p p values were provided in Supplementary Data 3). C Fluorescence microscopy of Ctr-AIDE dynamics in primary cells. Cdu1 (green, anti-FLAG), inclusions (Hsp60, red), and DNA (DAPI, blue). The white square in the left panel marks the region enlarged in the right panel. Scale bar = 10 µm. All experiments were replicated ≥3 times with consistent results. Source data are provided as a Source Data file.

Assessing Cdu1 Function Dynamics

Leveraging the Ctr-AIDE system, we first investigated the deubiquitinase function of Cdu1, previously implicated in countering autophagy signaling^36,37. We focused on the autophagy marker p62 to monitor autophagy-associated labeling at inclusions and bypass potential confounding effects of Ctr-AIDE-induced K48-linked ubiquitination (Supplementary Fig. 3). Depletion of Cdu1 triggered delayed recruitment of autophagy marker p62 to inclusions: the p62 signal on inclusions was minimal 1 h after 5-Ph-IAA treatment but markedly increased by 4 h (Fig. 4A). This phenotype was rapidly reversible with p62 recruitment dissipating within 1 h of 5-Ph-IAA washout (Fig. 4A). Quantification of p62-positive inclusions confirmed this phenotype: Cdu1 depletion triggered p62 accumulation, while re-expression led to loss of p62 staining (Fig. 4B).

The alternative text for this image may have been generated using AI.

A p62 recruitment to inclusions upon Cdu1 depletion. p62 (red, anti-SQSTM1), Cdu1 (green, anti-FLAG), DNA (DAPI, blue). The white square in the left panel marks the region enlarged in the right panel. Scale bar = 10 µm. B Quantification of p62-positive inclusions. Data represent mean percentage ± SD from 500 inclusions per condition across five independent experiments. Statistical significance was assessed by two-way ANOVA with Tukey’s test (****p p p C RT-qPCR of gapA (metabolic marker) and rrs (16S rRNA, control) at 48 hpi following timed 5-Ph-IAA treatment. DMSO treated samples, IFN-γ treated samples and Tn-Cdu1 strain served as controls³⁷. Data are normalized to DMSO-treated controls and shown as mean ± SD from three independent biological replicates. Statistical analysis by one-way ANOVA with Dunnett’s test (**p D Short-term depletion shows no effect. Quantification and statistical analysis were performed as in panel C.(*p E Timed Cdu1 depletion disrupts RB-to-EB redifferentiation. OmcB (EB marker). Cdu1 (anti-FLAG), OsTIR1(F74G) (anti-Myc), GAPDH (loading control). F Quantification of OmcB normalized to GAPDH (mean ± SD) from three replicates (one shown in E). Statistical significance assessed by one-way ANOVA with Dunnett’s test (**p p G Cdu1 loss reduces progeny-associated secondary infection burden. Lysates from primary infection (MOI = 1) were normalized by OmpA levels, then used to infect HeLa cells. H Quantification of OmpA abundance in secondary infections. OmpA was normalized to GAPDH (mean ± SD, n = 3; one replicate shown in G). Statistical analysis by one-way ANOVA with Dunnett’s test (**p p I Immunofluorescence microscopy of secondary infections. Inclusions (Hsp60, red) and DNA (DAPI, blue). Scale bar = 10 µm. All experiments were replicated ≥3 times with consistent results. Exact p values were provided in Supplementary Data 3. Source data are provided as a Source Data file.

To further evaluate Cdu1’s contribution to metabolic activity, and by extension overall fitness, we implemented two degradation strategies in the 2D PMRT cell model: (i) continuous degradation, in which 5-Ph-IAA treatment was initiated at 0, 8, 16, 24, 32, or 40 hpi and maintained until 48 hpi, and (ii) acute 8-hour pulses, initiated at staggered 8-hour intervals, and lasting 8 h each (Supplementary Fig. 8). Metabolic activity was monitored via RT-qPCR at 48 hpi for gapA (encoding glyceraldehyde-3-phosphate dehydrogenase), used here as a proxy for chlamydial metabolic activity under our assay conditions⁴⁵. Sustained depletion significantly reduced gapA expression when initiated before 24 hpi (Fig. 4C). In contrast, short-term 8-hour degradation had no significant effect, indicating that Cdu1’s metabolic role is only compromised by prolonged depletion (Fig. 4D). Consistent with previous reports^36,37, Cdu1 degradation did not impair metabolic activity in HeLa cells (Supplementary Fig. 9), by extension supporting the conclusion that Cdu1 is dispensable for chlamydial overall fitness in this cell type.

Cdu1 was also required for Ctr developmental progression³⁷. During mid-late chlamydial developmental, bacteria transition from the non-infectious reticulate body (RB) to the infectious, non-replicative elementary body (EB). Depletion before 24 hpi reduced levels of OmcB, a late-stage EB marker (Fig. 4E, F), coinciding with the onset of RB-to-EB differentiation⁴⁶. These findings highlight Cdu1’s critical role in initiating developmental transitions.

To assess the consequences of impaired EB formation, we evaluated progeny-associated infection burden using reinfection assays (Supplementary Fig. 10). Lysates from primary infected cells were used to infect fresh HeLa cells, and the extent of secondary infection was assessed by measuring OmpA levels at 24 hpi. OmpA (Major Outer Membrane Protein, MOMP) is highly expressed in EBs⁴⁷ and is critical for providing structural integrity to the chlamydial outer membrane. While OmpA abundance does not directly measure infectivity, it reflects bacterial load during the early stages of a new infection and thus serves as a surrogate readout of progeny-associated infection capacity. Cultures in which Cdu1 was depleted at or before 32 hpi yielded significantly reduced OmpA levels in secondary infections compared to untreated controls (Fig. 4G–I), consistent with impaired RB-to-EB differentiation in the primary culture. In contrast, no reduction in secondary infection burden was observed following acute 8-hour Cdu1 degradation pulses (Supplementary Fig. 11), reinforcing the requirement for sustained Cdu1 activity to support productive chlamydial development.

Inducible degradation and re-expression of IncA

Previous studies demonstrated that the C-terminus of IncA mediates inclusion fusion³⁹. To investigate the dynamics of IncA function as a key regulator of homotypic inclusion fusion^39,40, we applied the AIDE platform to this Inc protein. Using the same genome-integrated tagging strategy previously developed for Cdu1 (Supplementary Fig. 12), we generated Chlamydia strains expressing FLAG-tagged IncA or IncA-mAID-FLAG (Supplementary Fig. 1B).

Addition of 5-Ph-IAA induced rapid degradation of IncA-mAID within 1 hour, followed by re-expression within 30 min after 5-Ph-IAA washout. In addition, degradation required a functional E3 ligase OsTIR1(F74G), the mAID degron and addition of 5-Ph-IAA, since controls lacking any component or with dysfunctional E3 ligase retained stable IncA levels (Fig. 5A, B). Immunofluorescence microscopy also demonstrated the loss of inclusion membrane-localized IncA 1 hour after addition of 5-Ph-IAA and its restoration after 5-Ph-IAA washout (Fig. 5C).

The alternative text for this image may have been generated using AI.

A Ctr-AIDE-mediated regulation of IncA-mAID expression. Degradation and re-expression protocols were performed as described for Cdu1. IncA (anti-FLAG) and OsTIR1 mutants (TIR1-F74G denotes OsTIR1(F74G)-9×Myc, and TIR1-7K/10 K/74 G denotes OsTIR1(E7K/E10K/F74G)-9×Myc; anti-Myc) levels were monitored, with OmpA as a loading control and β-actin as a host cell control. B Quantification of IncA degradation. FLAG levels normalized to OmpA (mean ± SD) from three independent immunoblot replicates (one shown in A). Significance assessed by two-tailed paired t‑test (**p p p values were provided in Supplementary Data 3). C Immunofluorescence microscopy of IncA depletion dynamics. IncA (green, anti-FLAG), inclusions (Hsp60, red), and DNA (DAPI, blue). The white square in the left panel marks the region enlarged in the right panel. Scale bar = 10 µm. All experiments were repeated ≥3 times with consistent results. Source data are provided as a Source Data file.

Assessing IncA function dynamics

We next evaluated the functional consequences of IncA depletion. To remain consistent with previous study of IncA function⁴⁸, following experiments were performed at MOI = 1. Consistent with previous reports^39,40,48, IncA degradation by addition of 5-Ph-IAA from 0-24 hpi yielded a subset of infected cells containing multiple inclusions (Fig. 6A). This is consistent with the IncA’s essential role in mediating homotypic inclusion fusion.

The alternative text for this image may have been generated using AI.

A IncA degradation induces a multi-inclusion phenotype. HeLa cells expressing OsTIR1(F74G) infected with IncA-mAID strains (MOI = 1) were treated with 5-Ph-IAA (IncA degraded) or DMSO (control) from infection to 24 hpi. Degraded samples show cells containing multiple inclusions, whereas controls form single inclusions. IncA (green, anti-FLAG), inclusions (Hsp60, red), and DNA (DAPI, blue). Scale bar = 10 µm. B Schematic of IncA termination and re-expression experiments. C Immunofluorescence microscopy of inclusion phenotypes (MOI = 1) under Treatment 1-4 at the indicated time points. D Quantification of host cells containing multiple inclusions ( ≥ 2) upon Treatment 1-4. Data represents mean percentage ± SD from 200 cells counted per condition across five independent biological replicates. Significance assessed by two-way ANOVA with Tukey Multiple comparisons test (*p p values were provided in Supplementary Data 3). E Live-cell imaging reveals inclusion fission and fusion upon IncA degradation and re-expression. Representative time-lapse images from Treatments 1-4 illustrate inclusion fission during IncA degradation (Treatment 4) and inclusion fusion following IncA re-expression (Treatment 2). Continuous IncA depletion (Treatment 1) resulted in persistent inclusion fission, whereas continuous IncA expression (Treatment 3) yielded single inclusions. For treatment 1 and 3, the indicated hpi above marks the time at which the image was captured. For treatment 2 and 4, the indicated hpi on the left marks the time at which imaging was started, and the total imaging duration is indicated above the images. For each series, t0 and t_final denote the first and last frames, and inclusions were marked by white dashed line for clarity. Ctr (green, GFP), Scale bar = 10 µm. All experiments were replicated ≥3 times with consistent results. Source data are provided as a Source Data file.

Despite intensive research on the mechanism of inclusion fusion^40,48,49, it remains unclear whether IncA activity is continuously required to maintain fused inclusions or becomes dispensable once fusion is complete. Addressing this question has been challenging due to the long stability of membrane-integrated IncA, which limits the effectiveness of conventional silencing approaches. Continuous sRNA-mediated silencing of IncA expression⁵⁰ from 0 to 24 hpi effectively reduced IncA levels (Supplementary Fig. 13A), whereas silencing initiated at 24 hpi failed to deplete the protein, which remained stable for at least another 24 h under these conditions (Supplementary Fig. 13B, D). In contrast, AIDE enabled rapid depletion of membrane-integrated IncA within 1 h at 24 hpi (Supplementary Fig. 13C, D), providing a powerful tool to directly assess the temporal requirements of IncA function. To determine whether IncA’s function is static or dynamic, we implemented time-resolved degradation schemes. We compared two conditions: (i) Expression pause: IncA expressed from 0 to 24 hpi, then degraded until 40 hpi (Treatment 4). (ii) Rescue: IncA degraded from 0 to 24 hpi, then re-expressed via 5-Ph-IAA washout until 40 hpi (Treatment 2). Controls included continuous IncA expression (0-40 hpi, Treatment 3) or degradation (0-40 hpi, Treatment 1) (Fig. 6B).

Control treatments established the expected inclusion phenotypes: continuous IncA expression was associated with predominantly single inclusions per infected cell (Treatment 3), whereas sustained IncA degradation resulted in ~25% of infected cells containing multiple inclusions from 24 to 40 hpi, likely representing the subset of cells in which IncA-dependent homotypic fusion would normally occur (Treatment 1) (Fig. 6C, D). Quantitative analysis further revealed that initiating IncA degradation at 24 hpi (“Expression pause,” Treatment 4) increased the fraction of infected cells containing ≥2 inclusions compared to the continuous-expression control (Fig. 6C, D). In contrast, restoring IncA after washout at 24 hpi (“Rescue,” Treatment 2) reduced the proportion of cells with multiple inclusions toward control levels (Fig. 6C, D).

Because the penetrance of multi-inclusion phenotypes can depend on infectious input^51,52, we examined the effect of IncA depletion and re-expression across different MOIs. Using the same Treatment 1-4 scheme, we quantified the number of inclusions per infected cell at 24 and 40 hpi at MOI = 0.1, 1, and 10 (Supplementary Fig. 14). At MOI = 0.1, no significant differences in inclusion number were observed across treatments. In contrast, at MOI = 1 and MOI = 10, IncA depletion at 24 hpi (treatment 4) increased the number of inclusions per infected cell at 40 hpi, whereas IncA re-expression at 24 hpi (treatment 2) shifted this distribution back toward lower inclusion numbers at 40 hpi (Supplementary Fig. 14). These results are consistent with previous studies showing that IncA-related multiple-inclusion phenotypes become more apparent at higher MOI^51,52.

To determine whether these changes in inclusion number per cell reflect underlying fusion and fission dynamics, we performed live-cell imaging under the same treatment conditions at MOI = 1. Live-cell imaging directly captured these dynamics, showing inclusion fission upon rapid IncA depletion started at 24 hpi and inclusion fusion following IncA re-expression started at 24 hpi (Fig. 6E, Supplementary Movies 1–12). These results provide direct mechanistic evidence that IncA is not only required for the establishment of homotypic inclusion fusion, but also for maintenance of the fused-inclusion state, as loss of IncA leads to inclusion fission.

To rule out that the multi-inclusion phenotype is an artifact of the AIDE degradation system rather than a consequence of IncA loss, we performed a control experiment in which Cdu1 was degraded continuously from infection through 48 hpi and inclusion morphology was assessed. In contrast to IncA depletion, prolonged Cdu1 degradation did not increase the fraction of infected cells containing multiple inclusions (Supplementary Fig. 15), indicating that multi-inclusion phenotype is not a general outcome of AIDE activation and OsTIR1(F74G) recruitment to the inclusion and supporting the conclusion that the phenotype is attributable to IncA depletion.

To align with the Cdu1 functional analyses, we also tested the effects of IncA manipulation on chlamydial metabolism, developmental progression, and progeny-associated secondary infection burden using the same treatment scheme shown in Fig. 6B.

In primary cells, IncA degradation triggered multi-inclusion phenotype (Supplementary Fig. 16A). In the 2D PMRT model, IncA depletion reduced chlamydial metabolic activity (Supplementary Fig. 16B, left, treatment 1, 4 vs treatment 3). Notably, re-expression of IncA from 24-40 hpi did not fully restore metabolic activity (Supplementary Fig. 16B, left, treatment 2 vs treatment 3), suggesting that sustained IncA activity is required to maintain normal metabolic fitness in primary cells. In contrast, IncA degradation in HeLa cells did not measurably affect metabolic activity (Supplementary Fig. 16B, right).

In primary cells, prolonged IncA depletion from early infection through 40 hpi (treatment 1) impaired developmental progression, as indicated by reduced OmcB accumulation (Supplementary Fig. 16C, D; treatment 1 vs treatment 3), and was accompanied by reduced progeny-associated secondary infection burden in reinfection assays (Supplementary Fig. 16E, F; treatment 1 vs treatment 3). Restricting IncA depletion to 24-40 hpi caused a modest, non-significant reduction in OmcB levels (Supplementary Fig. 16C, D; treatment 4 vs treatment 3), yet this was accompanied by a significant reduction in progeny-associated secondary infection burden (Supplementary Fig. 16E, F; treatment 4 vs treatment 3), suggesting that even a subtle impairment in late-stage EB formation can translate into a measurable reduction in progeny-associated secondary infection burden. Consistent with these findings, IncA re-expression at 24 hpi did not fully restore developmental progression and secondary infection burden, mirroring the incomplete recovery of inclusion fusion (Supplementary Fig. 16C–F; treatment 2 vs treatment 3). In HeLa cells, by contrast, IncA degradation did not significantly alter OmcB levels or progeny-associated secondary infection burden (Supplementary Fig. 17A–D).

Together, these results indicate that IncA is particularly important for productive infection in primary cells, where loss of IncA disrupts inclusion fusion and may increase susceptibility to host restriction mechanisms, whereas IncA is largely dispensable for chlamydial metabolic activity and developmental progression in HeLa cells.