The Chlamydia trachomatis type III effector TarP coordinates a functional collaboration between the actin nucleators Formin 1 and Arp2/3 during invasion

The obligate intracellular pathogen Chlamydia trachomatis manipulates the host actin cytoskeleton to assemble actin-rich structures that drive pathogen entry. This actin remodeling event exhibits relatively rapid dynamics that, through quantitative live-cell imaging, was revealed to consist of three phases – a fast recruitment phase which abruptly transitions to a fast turnover phase before resolving into a slow turnover of actin that indicates the end of actin remodeling. Here, we investigate Chlamydia invasion in the context of actin dynamics. Efficient invasion is associated with robust actin remodeling kinetics that results from a collaborative functional interaction between two different classes of actin nucleators – formins, including formin 1 and the diaphanous-related formins mDia1 and mDia2, and the Arp2/3 complex. Recruitment of these nucleators requires the presence of the chlamydial type III effector TarP, which enables the respective nucleating activities of formin and Arp2/3 to collaboratively generate a robust actin network. A collaborative model is supported by the observation that co-inhibition of Fmm1 and Arp2/3 further reduced both actin dynamics and invasion efficiency than either treatment alone. Furthermore, inhibition of recruitment of Fmn1 and/or Arp2/3 by deleting TarP was sufficient to similarly attenuated actin kinetics and invasion efficiency, supporting a model wherein TarP is the major contributor to robust actin remodeling via its recruitment of the two classes of actin nucleators. At the population level, the kinetics of recruitment and turnover of actin and its nucleators were linked. However, a more detailed analysis of the data at the level of individual elementary bodies showed significant variation and a lack of correlation between the kinetics of recruitment and turnover, suggesting that accessory factors variably modify actin kinetics at individual entry sites. In summary, efficient chlamydial invasion requires a specific profile of actin dynamics which are coordinated by TarP-dependent recruitment of two classes of actin nucleators. Author Summary The obligate intracellular pathogen Chlamydia trachomatis relies upon manipulation of the host actin cytoskeleton to drive its entry into host cells, such that impairment of actin dynamics attenuates Chlamydia invasion. Collaboration between two classes of actin nucleators, formin and Arp2/3, are known to enhance actin recruitment and turnover; we found that recruitment of both proteins to the signaling complex established by the type III secreted effector, TarP, was important for pathogen internalization. Furthermore, Formin 1 and Arp2/3 are co-recruited to sites of entry, and pharmacological inhibition of either actin nucleator impaired recruitment of the other, indicating a functional cooperation between branched and filamentous actin nucleation within pathogen entry sites. Disruption of this cooperation negatively impacted both actin dynamics and Chlamydia internalization, indicating that TarP-dependent entry of Chlamydia into non-phagocytic cells operates through the recruitment and activation of Arp2/3 and Formin 1. Finally, kinetic analysis of actin recruitment and turnover revealed that these processes were independently regulated, in addition to implicating the presence of local factors that fine-tune actin dynamics and subsequent invasion.


18
The obligate intracellular pathogen Chlamydia trachomatis manipulates the host actin cytoskeleton to 19 assemble actin-rich structures that drive pathogen entry. This actin remodeling event exhibits relatively 20 rapid dynamics that, through quantitative live-cell imaging, was revealed to consist of three phases -a 21 fast recruitment phase which abruptly transitions to a fast turnover phase before resolving into a slow burgdorferi (21)(22)(23). Likewise, invading C. trachomatis is known to interact with hypertrophic microvillar 82 structures that are enriched with filamentous actin (12), raising the possibility that Chlamydia also utilizes 83 formins during invasion.

84
Arp2/3 is a well-known nucleator of actin filaments that is recruited during chlamydial invasion 85 following the activation of WAVE2 and the Rho-GTPase Rac1 by the C. trachomatis effector TarP (9,24).    Table 1), which is comparable to the reduction observed in the recruitment phase of actin.

213
Moreover, inhibition of formin or Arp2/3 also altered the turnover profile of actin, generating a narrower 214 range of turnover rates (Table 1). Given that recruitment and turnover are comparably altered by 215 inhibition of formin or Arp2/3, our data imply a link between these phases such that rapid/slowed 216 recruitment is paired with rapid/slowed turnover. This is also true for when formin and Arp2/3 are 217 simultaneously inhibited, which reduces the turnover rate of actin by over 10-fold (CK+SMI=0.005 218 fold/sec) compared to mock treatment. In contrast, slow turnover rates exhibited only minor variations 219 between conditions, even in circumstances where both formin and Arp2/3 are inhibited (Fig. 1E, right).

220
To confirm that comparisons derived from the violin plots are reflective of the overall kinetics for each 221 recruitment condition, we performed a linear regression analysis on the mean rate of recruitment and 222 turnover for each phase (Fig. S3). We noted that the trends observed in violin plots (Fig. 1D,E) were also 223 observed in the averaged rate slopes for all conditions tested ( Table 2). Together, this indicates that collaboration between the activities of Fmn1 and Arp3 not only 294 enhance the rate of actin recruitment (Fig. 1D), but also reciprocally enhance their respective recruitment 295 rates. Although the recruitment profiles of Fmn1 and Arp3 were largely comparable, we found that their 296 fast turnover profiles exhibited substantial differences. In particular, we observed that mock-treated 297 Fmn1 achieved a median fast turnover rate that was over twofold higher than that of mock-treated Arp3  (Fig. S4B). In sum, these data suggest that Fmn1 and Arp2/3 collaboration is necessary for 313 rapid turnover of Fmn1, but not Arp2/3. While we observed some slight differences in the slow turnover 314 phases of Fmn1 and Arp3 (Fig. 2F), we noted that the variability of these data are quite high. As such, it is 315 difficult to give a precise account of how reciprocal inhibition affects the residual turnover of Fmn1 and 316 Arp3, although the miniscule turnover rates within this phase suggest that its contribution to the overall 317 turnover of Fmn1 and Arp2/3 is minor. Finally, we noted that the recruitment and turnover kinetics of 318 mDia1 and mDia2 were comparable to Fmn1 both in mock-and CK666-treated groups (Fig. S5, Table S1),

319
indicating that all formin species are regulated comparably during invasion. Altogether, we conclude that 320 kinetics of recruitment and turnover of Fmn1 and DRFs are enhanced by collaboration with Arp2/3, and 321 reflect the kinetics observed for actin recruitment and turnover (Fig. 1D,E) unresponsive to the subsequent application of inhibitors to formin or Arp2/3 (Fig. 3B). In agreement with 344 previous findings, TarP exhibited a substantial impairment to invasion efficiency relative to cis-TarP (40).

345
Contrary to our expectations, inhibition of formin or Arp2/3 resulted in a further reduction in pathogen 346 internalization, however, invasion efficiency was not substantially worsened by simultaneous inhibition.

347
This suggests that while the invasion-associated activities of formin and Arp2/3 are not entirely TarP 348 dependent, deletion of TarP renders Chlamydia invasion more sensitive to formin and/or Arp2/3 349 inhibitors.

350
To more precisely evaluate the effect of TarP deletion on actin recruitment, we tracked the 351 accumulation of mRuby-LifeAct at Chlamydia entry sites and analyzed the kinetics of actin recruitment 352 and turnover (Fig. 3 C-E). We observed that cis-TarP robustly recruited actin, resulting in a 3-fold increase 353 in actin accumulation compared to TarP (Fig. 3C). We noted that cis-TarP featured a median actin 354 recruitment rate that is comparable to wild-type C. trachomatis (cis-TarP=0.233 fold/sec vs. wild-355 type=0.285 fold/sec) (Fig. 1D), while loss of TarP reduced the rate of actin recruitment by 4-fold 356 (TarP=0.054 fold/sec) (Fig. 3D, Table 3). Likewise, TarP deletion slowed the rate of actin turnover by 357 nearly 3-fold (cis-TarP=0.117 fold/sec vs. TarP=0.043 fold/sec) (Fig. 3E), indicating that TarP contributes 358 not only to the rapid recruitment of actin, but also assists in facilitating rapid turnover. We did not observe 359 any correlation between the rate of recruitment and turnover in cis-TarP (R 2 =0.115) and weak correlation 360 in TarP (R 2 =0.287), suggesting that TarP alone is insufficient to explain the independent regulation of 361 actin recruitment and turnover (Fig. S6A). However, TarP deletion also alters the slow turnover phase of 362 actin (Fig. 3E), eliminating the boundary between rapid and lingering actin turnover that is readily   Fig. 4A-C). While cis-TarP robustly recruited both Fmn1 and Arp3, TarP experienced attenuated 377 Arp3 recruitment and failed to recruit Fmn1 altogether (Fig. 4A). Quantitative imaging of Fmn1 and Arp3 378 recruitment revealed that Arp3 recruitment in TarP was 2-fold less robust than cis-TarP, while only 379 recording trace signals of Fmn1 recruitment in a TarP background (Fig. 4B,C) (Fig. 4E, Table 4). By comparison, reciprocal 388 inhibition of Fmn1 or Arp3 slowed the recruitment of either nucleator by just over 2-fold (Fig. 2E).

389
Additionally, we noted that the rate of Fmn1 recruitment in a TarP background was nearly negligible; 390 barring outliers, the median recruitment rate clustered around 0.0093 fold/sec, which is consistent with 391 our earlier assessment that Fmn1 is likely not recruited in the absence of TarP (Fig. 4A,B). In sum, these 392 data indicate that TarP contributes to the recruitment of Fmn1 and Arp3, and that loss of TarP significantly 393 attenuates the recruitment of both nucleators. Arp3=0.023 fold/sec) (Fig. 4E, Table 4), suggesting that TarP also contributes to rapid turnover of Fmn1 402 and Arp3. In contrast, the slow turnover rates of Fmn1 and Arp3 were mostly unaffected by TarP deletion, 403 suggesting that TarP predominantly affects the rapid turnover phase of formin and Arp2/3 (Fig. 4E). Since 404 the turnover of Fmn1 and Arp2/3 (Fig. 2F) is linked to actin turnover (Fig. 1E), the role of TarP as a 405 determinant of Fmn1 and Arp2/3 turnover kinetics is likely related to its effects on the turnover of the 406 actin network assembled post-recruitment (Fig. 3), and to which these nucleators bind. This is further 407 supported by the fact that recruitment and turnover rates are not correlated in Fmn1 (R 2 =0.111), Arp3 408 (R 2 <0.001), or actin (R 2 =0.115) (Fig. S6)

459
In addition to rapid actin recruitment, we also observed a two-phase actin turnover process 460 starting with a brief fast turnover phase (100-180 sec) followed by a prolonged slow turnover phase (400- While our data impress the importance of TarP in promoting the recruitment and turnover of 488 Arp2/3, which occurs via signaling through the WAVE2 complex, we also observed that a portion of Arp2/3 489 was recruited in a TarP-independent manner. Given recent evidence demonstrating that the Chlamydia 490 effector TmeA also activates Arp2/3 during invasion by signaling through N-WASP it is likely that the 491 residual recruitment of actin and Arp2/3 in ΔTarP strains is due, at least in part, to the activity of TmeA 492 (Fig. 5A) (41,42). A direct comparison between WT, ΔTarP, ΔTmeA, ΔTarP/ ΔTmeA strains with regards to 493 dynamics of actin, Fmn1, and Arp2/3 at the invasion site would be required to assess accurately their 494 relative contributions. However, some inferences can be drawn from the data reported here. We 495 observed that the predominant fraction of Arp2/3 recruitment was linked to TarP (Fig. 4C), indicating that 496 the contribution of TmeA in this process is auxiliary. Furthermore, invasion of TarP-deficient mutants that 497 retained wild type TmeA was associated with a radically different actin kinetics, i.e. a slower rate of 498 recruitment and an extremely slow turnover (Fig. 3B-E), which would be expected to be associated with TarP signaling contributes to the assembly of both branched and filamentous actin (Fig. 5B)

650
Tabulated statistical analyses of violin plots were generated using base R statistics and the moments 651 package (version 0.14, https://cran.r-project.org/web/packages/moments/index.html) in rStudio and shown previously (Fig. 1C) were individually divided into recruitment, fast turnover and slow turnover 860 phases (Fig. S2). (D) Individual rates of recruitment were plotted on a violin plot with inset boxplot,