Intracellular Cryptococcus neoformans disrupts the transcriptome profile of M1- and M2-polarized host macrophages

Macrophages serve as a first line of defense against infection with the facultative intracellular pathogen, Cryptococcus neoformans (Cn). However, the ability of these innate phagocytic cells to destroy ingested Cn is strongly influenced by polarization state with classically (M1) activated macrophages better able to control cryptococcal infections than alternatively (M2) activated cells. While earlier studies have demonstrated that intracellular Cn minimally affects the expression of M1 and M2 markers, the impact on the broader transcriptome associated with these states remains unclear. To investigate this, an in vitro cell culture model of intracellular infection together with RNA sequencing-based transcriptome profiling was used to measure the impact of Cn infection on gene expression in both polarization states. The gene expression profile of both M1 and M2 cells was extensively altered to become more like naive (M0) macrophages. Gene ontology analysis suggested that this involved changes in the activity of the Janus kinase-signal transducers and activators of transcription (JAK-STAT), p53, and nuclear factor-κB (NF-κB) pathways. Analyses of the principle polarization markers at the protein-level also revealed discrepancies between the RNA- and protein-level responses. In contrast to earlier studies, intracellular Cn was found to increase protein levels of the M1 marker iNos. In addition, common gene expression changes were identified that occurred post-Cn infection, independent of polarization state. This included upregulation of the transcriptional co-regulator Cited1, which was also apparent at the protein level in M1-polarized macrophages. These changes constitute a transcriptional signature of macrophage Cn infection and provide new insights into how Cn impacts gene expression and the phenotype of host phagocytes.

. 250 Cn, GS is goat serum. Error is represented as S.E. Statistical differences between samples 251 were appraised by one-way analysis of variance (ANOVA) followed by a Tukey's multiple 252 comparison test. Statistical significance is indicated as follows: *, p < 0.05; **, p< 0.01.
253 With the exception of (B-D), data is from three biological repeats.

255
As macrophages will typically repolarize from an M1 to an M2 state rather than 262 from M1 to M2 resulted in >100 fold increase in Arg-1 protein levels ( Fig 1E).
263 Collectively, these data suggested that RAW 264.7 cells could be used to model both the 264 M1 and M2 state with the caveat that the M2 phenotype could not be reached directly 265 from the naive state but via repolarization from M1 ( Fig 1E).

266
Given that Cn can replicate within host cell phagolysosomes and can escape into  296 macrophages ( Fig 1I). We found that a concentration of 2× 18B7 (20 μg for 1.5 x 10 6 297 cells) was sufficient to routinely obtain infection efficiencies of ~80%, as determined at 298 the time of macrophage harvest for mRNA extraction. While it is common to add LPS to 299 macrophages to promote phagocytosis, we found that this did not affect the percentage 300 of Cn-infected macrophages ( Fig 1I). Furthermore, as co-infections with gram-negative Based on the data described in Figure 1, we developed a strategy to produce Cn-311 infected M1 and M2-polarized RAW 264.7 cells (Fig 2A). In brief, this involved M1-      (Table 3 and    523 Given that culture conditions were tightly controlled throughout our experiments with 524 growth medium and cytokines replaced at regular intervals, we believe that these 525 changes are genuine and not simply caused by relaxation back to an M0 state, as might 526 happen on withdrawal of polarizing stimuli. Additionally, we identified a common set of 527 genes that were affected in a similar fashion regardless of host cell polarization state 528 and constitute a transcriptional signature of intracellular Cn infection (Table 3 and 4).

529
The effect of intracellular Cn on the host macrophage transcriptome has been

544
Our analysis also allowed us to identify common and differing cellular processes 545 affected by Cn infection in M1 and M2 cells. The most notable common effect was the 546 reversal of transcriptome changes associated with phagocytosis and lysosomal function 547 (Fig 3A and 4A (Table 3 and 4).
577 This latter group included the transcriptional regulators Cited1, Hsf3, and Jarid2, the 578 cytokine Ccl2, and the anti-apoptotic factor Bcl2a1a. Of these, Cited1, which was 579 previously known as melanocyte-specific protein 1 (Msg1), showed the largest fold 580 change and was also upregulated at the protein level post-Cn infection (Fig 5). 592 This is the subject of ongoing investigations in our lab and our findings will be described 593 in detail as part of a future publication.

604
Finally, our study also highlights the importance of carefully controlling nutrient 605 levels in in vitro culture systems used to study the Cn:macrophage interaction, especially 606 if live rather than heat-killed Cn are utilized and experiments last more than a few 607 hours. In preliminary experiments that preceded this study, we detected a significant 608 reduction in the expression of the glucose-responsive gene TXNIP at the transcript and 609 protein level ( Fig 1G). As TXNIP is a regulator of p53, NF-κB, and other pathways 610 associated with the control of macrophage polarization, glucose depletion may impact 611 macrophage phenotype and result in spurious gene expression changes that do not