Karyopherin Alpha Proteins Regulate Oligodendrocyte Differentiation

Proper regulation of the coordinated transcriptional program that drives oligodendrocyte (OL) differentiation is essential for central nervous system myelin formation and repair. Nuclear import, mediated in part by a group of karyopherin alpha (Kpna) proteins, regulates transcription factor access to the genome. Understanding how canonical nuclear import functions to control genomic access in OL differentiation may aid in the creation of novel therapeutics to stimulate myelination and remyelination. Here, we show that members of the Kpna family regulate OL differentiation, and may play distinct roles downstream of different pro-myelinating stimuli. Multiple family members are expressed in OLs, and their pharmacologic inactivation dose-dependently decreases the rate of differentiation. Additionally, upon differentiation, the three major Kpna subtypes (P/α2, Q/α3, S/α1) display differential responses to the pro-myelinating cues T3 and CNTF. Most notably, the Q/α3 karyopherin Kpna4 is strongly upregulated by CNTF treatment both compared with T3 treatment and other Kpna responses. Kpna4 inactivation results in inhibition of CNTF-induced OL differentiation, in the absence of changes in proliferation or viability. Collectively, these findings suggest that canonical nuclear import is an integral component of OL differentiation, and that specific Kpnas may serve vital and distinct functions downstream of different pro-myelinating cues.


Introduction
Precise transcriptional regulation of oligodendrocyte (OL) differentiation, influenced by extrinsic stimuli, is required for OL development and subsequent central nervous system (CNS) myelination [1]. As myelin is essential for efficient action potential transmission in the CNS [2], failure of myelination or loss of myelin (demyelination) results in symptoms of white matter diseases such as leukodystrophies and multiple sclerosis (MS) [3,4]. Myelin repair (remyelination) restores conduction and brings functional recovery in diseases such as MS, but is inefficient, and often fails in the disease's later stages [1]. Disruption of white matter formation and repair is characterized by failure of oligodendrocyte progenitors (OLPs) to mature into myelinating OL [1]. Currently, treatments to promote OL differentiation, and thus myelin a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Gapdh and Actin was 70,077 ± 397.2 and 153,795 ± 2222.0 a.u. respectively). While NanoString analysis is impacted by probe binding efficiency, these data do suggest that Kpna2 displays significantly higher expression than other paralogs in specified OLP.

Canonical transport is an important regulatory mechanism in oligodendrocyte differentiation
With widespread expression detected, we hypothesized that canonical import was essential to OL development, particularly during differentiation, a period when a wide range of transcription factors undergo coordinated entry into the nucleus. To test this hypothesis, we differentiated mouse OLP using T3 in the presence of increasing concentrations of importazole (IPZ), a well-characterized pharmacological inhibitor of Kpna/b-mediated nuclear import [29]. Importantly, inactivation of all canonical import strongly inhibited OLP differentiation as evidenced by a dose-dependent inhibition of the normal increase in expression of maturation and myelin markers (CNP and MBP) detected via quantitative PCR (qPCR) (Fig 2B). Moreover, although concentrations of IPZ up to 2μM produced dose-dependent inhibition of differentiation in the absence of any impact on viability, the highest dose of IPZ used (20μM) produced an increase in apoptosis, observed as a rise in the percentage of cleaved-caspase 3 (CC3) expressing cells (Fig 2C and 2D). Thus, while the effects of decreased expression of maturation markers at the highest dose of IPZ may relate in part to the loss of cells due to programmed death, lower doses produced reductions in maturation marker expression without changes in viability ( Fig  2B). Indeed, extension of differentiation out to 72h with T3 in the presence of 2μM IPZ confirmed a sustained reduction in maturation (Fig 2E-2I). As such, these results suggest that canonical nuclear import is an important regulatory mechanism in OL differentiation, and may also play a broader role in viability.
Karyopherin alphas exhibit differential transcriptional profiles during oligodendrocyte differentiation. Since Kpna/b-mediated nuclear import regulates OL differentiation, we wondered how expression of each paralog changes in the differentiation program. During differentiation in other lineages, a phenomenon known as isotype switching has been observed, whereby Kpna isoforms swap expression predominance in order to enact a switch from Kpna gene expression in OLP exposed only to proliferative factors (PDGFAA, 10 ng/ml and bFGF, 20 ng/ml) for 24h. Expression was quantified from isolated total RNA with NanoString nCounter Gene Expression Assay. A panel of probes was designed for target genes and normalized to housekeeping genes Alas1, Ppia, Gapdh, Actb, and Rps11. Following assay completion, raw data was normalized and analyzed using nSolverTM software. Kpna2 is the highest expressing Impα in OLP. Kpna7 is not expressed in oligodendrocytes. (B-D) Results from primary OLP pre-treated 1h with increasing concentrations the karyopherin inhibitor importazole (IPZ) or DMSO, then differentiated with T3 (40ng/ml) in the presence of IPZ/DMSO for 9h and harvested for qPCR (B) or immunocytochemistry (C). (B) Failure of karyopherin functioning in IPZ-treated cultures results in dose-dependent inhibition of differentiation as measured by a decrease in the expression of the OL maturation markers CNP or MBP in qPCR. (C,D) While, at lower doses of IPZ, inhibition of differentiation is seen in the absence of any impact on viability, higher concentrations of IPZ eventually produce apoptosis, quantified as the percentage of DAPI cells that are cleaved-caspse 3 (CC3) positive. (E-I) Extension of differentiation out to 72h with T3 in the presence of 2μM IPZ confirms this reduction in maturation. Results are from primary OLP pretreated 1h with 2μM IPZ or DMSO, then differentiated with T3 (40ng/ml) in the presence of IPZ/DMSO for up to 72h and harvested for qPCR maintenance of the progenitor state towards progression of differentiation [11,12]. However, this is not a universal mechanism used by all cells [13,30]. Given the high expression in OLP of one alpha isoform (Kpna2) compared to the rest, we hypothesized that the potential for differential changes in Kpna isotype expression may exist in these cells. Additionally, we suspected that different isotypes would be selectively altered based on the pro-myelinating cue used, since different Kpnas could carry a different set of transcription factors (i.e. different cargo) as part of downstream mechanisms specific to each cue.
To test this hypothesis, we differentiated OLP using either T3 (60 ng/ml) or CNTF (100 ng/ ml) and harvested for RNA at 24h intervals through 72h. This latter time-point corresponds to the emergence of mature and arborized OL positive for the early maturation marker, O4, and the later myelin protein, MBP regardless of the differentiating factor used (Fig 3A). Equivalent maturation is also observed at earlier time points, as demonstrated by qPCR confirming similar increases in RNA for the myelin components CNP and MBP in response to T3 and CNTF by 24h of differentiation (Fig 3B). Kpna paralog expression was thus assessed from 0-72h of treatment with T3 or CNTF, at 24h intervals.
Notably, while the ranking of expression levels in OLP was maintained throughout differentiation and seen in more mature OL at 72h (see S1 Table), these studies showed that each individual Kpna underwent significant changes in expression during the differentiation program. Differential changes in expression of Kpna paralogs were detectable from 24h of differentiation (Fig 3C-3F). We found that Kpna2 decreased by approximately 70% of its OLP value in response to both T3 and CNTF. In contrast, the other paralogs all increased their expression during differentiation regardless of the initiating stimulus, though with different magnitudes (Fig 3C and 3D). Thus, the predominant response of Kpnas is to increase expression of family members with differentiation. We compared these results to Kpnb1, the binding partner for Kpna within the Kpna/b heterodimer. Unlike the significant increases seen in the expression of Impαs, Kpnb1 showed more stable levels throughout differentiation, albeit with slight fluctuations in expression compared to its OLP levels (Fig 3G).
Intriguingly, these data further revealed that Kpna subtypes respond differently to different pro-myelinating cues, as evidenced by differential responses to T3 versus CNTF. While Kpna2 (Subtype P/α2) decreased no matter the extrinsic factor, Kpna4 (Subtype Q/α3) showed a greater response to CNTF compared to other members (Fig 3C-3F and 3H, S1 Table). Conversely, Kpna1 and Kpna6 (Subtype S/α1) displayed greater fold changes in response to T3 compared to members of subtypes P/α2 and Q/α3, as well as responsiveness to CNTF (Fig 3F and S1 Table). These findings were confirmed via immunoblotting for representative family members (Kpna1, Kpna2, and Kpna4) in response to CNTF (Fig 3I and 3J). Thus, each of the three major Kpna subtypes (P/α2, Q/α3, S/α1) showed a differential response to T3 and CNTF: P/α2 decreased expression to both, Q/α3 increased more substantially to CNTF, and S/ α1 had greater responsiveness to T3 than other subtypes.

Effects of inactivation of Kpna4 is more severe in the presence of CNTF differentiation
Since we have previously shown that Kpna1 is an important regulator of OL differentiation, our attention was drawn to the data for Kpna4 in the above studies (Fig 3H-3J), as it showed the most striking difference in response to pro-myelinating cues, as well as one of the greatest magnitude changes associated with differentiation, which additionally was CNTF-specific. To determine in vivo relevance of our NanoString findings for Kpna4, we assessed its expression in developing spinal cord, a well-characterized system for studying myelination in vivo [1]. In the spinal cord myelination proceeds with a rostral to caudal gradient (especially ventrally) Confocal images of mature OL treated with either T3 (60ng/ml) or CNTF (100ng/ml) for 72h. Cells were immunostained for the earlier maturation marker, O4 (green), and the later myelin protein, MBP (red), a marker of mature OL. (B) Results from qPCR of fold change expression for the maturation markers CNP and MBP at 24h following T3 (left) or CNTF (right) treatment, compared to 0h. (C-H) Kpna gene expression data from OL lineage for 72h treatment with either T3 or CNTF, analyzed at 24h intervals. Transcripts were quantified from isolated total RNA using NanoString nCounter Gene Expression Assay. A commercially available panel of probes for target genes was normalized to housekeeping genes Alas1, Ppia, Gapdh, Actb, and Rps11. Following assay completion, raw data was analyzed using nSolver [31]. Conversely in the brain it progresses from caudal sites to rostral ones. CNS myelination is extensive by postnatal day 14 (P14). Confocal imaging of P1 and P14 thoracolumbar spinal cords demonstrated widespread immunoreactivity for Kpna4 in OL expressing the lineage marker Olig2 and the maturation marker APC in the more mature ventrolateral funiculi, but not by immature Olig2 + APCcells (Fig 4A and 4C). These cells exhibited both nuclear and cytoplasmic Kpna4 expression, compatible with its function in shuttling cargo between both regions (Fig 4B).
Interestingly, Kpna4 inactivation in primary cultures implicated this import factor as an important regulator of the OL differentiation program downstream of CNTF-induced signaling. To examine its involvement, we inactivated Kpna4 via nucleofection of an siRNA targeted specifically to Kpna4 and compared it to a nontargeting (siNT) control. To directly compare responses of Kpna4-deficient cells and controls to CNTF vs. T3 in parallel, primary cultures were nucleofected with siRNA for Kpna4 or nontargeting (NT) control, then treated with either CNTF or T3. Kpna4 inactivation was efficient and selective-other isoforms did not decrease with silencing (Fig 4D). Kpna4 silencing did not alter active proliferation of OLPs as assessed by immunocytochemistry for Ki67 followed by confocal imaging and quantitation of %Ki67 + (%Ki67/DAPI) cells. The percentage of Ki67 + OLP was unchanged between groups in primary cultures (Fig 4E and 4F). Following exposure of OLP to either T3 or CNTF, using confocal microscopy we determined the percentage of OL expressing the maturation markers O4 and MBP. Compatible with its expression patterns in our NanoString analysis (see Fig 3), differentiation was strongly inhibited by Kpna4 inactivation in CNTF-treated cultures (Fig  4G-4K). There was a strong reduction in the accumulation of O4-expressing cells in CNTFtreated cultures deficient in Kpna4 versus nontargeting controls (Fig 4H). Similarly, the induction of MBP expression was also substantially decreased in CNTF-treated cultures in which Kpna4 was silenced (Fig 4I). Cells also displayed a less complex branching morphology, as measured by fractal analysis (Fig 4J). Conversely, we observed only a slight trending decrease in the percentage of MBP expressing cells in T3-treated cultures in which Kpna4 was inactivated (p = 0.14) (Fig 4I), with no significant difference in arborization (Fig 4K), in line with the more minor increase in Kpna4 expression observed in response to T3 treatment. This suggests that Kpna4 inactivation may slightly delay differentiation to the more mature stage in response to T3, whereas in CNTF-treated cultures its loss causes a more severe inhibition or delay of differentiation. Importantly, these changes seen in differentiation were not due to cell death, as both cell number and viability (measured as the percentage of apoptotic cells displaying cleaved-caspase 3) remained unaltered after Kpna4 inactivation, compared to siNT control, in both T3-and CNTF-treated cultures (Fig 4L and 4M).
software before being subjected to statistical analysis. (C-F) Expression fold change in response to T3 (60ng/ml) or CNTF (100ng/ml) at 24h intervals, derived from NanoString analysis. Results demonstrate differential changes in expression of Kpna isotypes during differentiation. Kpna2 expression strongly decreases, whereas all other isoforms increase. The three Kpna subtypes (P/α2, Q/α3 and S/α1) also respond differently to the extrinsic factor used. While Kpna2 (Subtype P/α2) decreases no matter the cue, Kpna4 (Subtype Q/α3) shows greater changes in expression in response to CNTF, and Kpna1 and Kpna6 (Subtype S/α1) display greater fold changes in response to T3 than other subtypes, in addition to responsiveness to CNTF. (G) Expression fold change derived from NanoString analysis shows that Kpnb1 displays relatively stable expression throughout the course of OL differentiation. (H) Expression fold change derived from NanoString analysis demonstrates differential responses of Kpna4 to CNTF versus T3 treatment. While expression increases slightly in response to T3, the response to CNTF is greater at all time points beyond 0h. (I,J) Gene expression changes resulted in corresponding alterations in protein levels in response to CNTF. OLP treated for up to 72h with CNTF (100ng/ml) were harvested and immunoblotted for Kpna1, Kpna2, and Kpna4, with Actin used as a loading control (I). (J) Accompanying densitometry plots for Kpna4 were calculated from the ratio of Kpna4/Actin pixel intensity and displayed as fold change from 0h of CNTF treatment. Data are mean ± S.E.M. and representative of 3 independent cultures. Statistics, (B) Student's t-test, (C-H) Two-way ANOVA plus Bonferroni post-test, (J) One-way ANOVA plus Bonferroni post-test, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Statistics for (E-F) are in S2 Table. Scalebar: (A) 20μm. Individual data values for (C-H) are in S1 Table and S1 Data. Individual data values for (B,J) are in S1 data.

Discussion
This study shows that Kpnas are important regulators of OL differentiation, and suggests that different members of the family may play distinct roles downstream of different pro-myelinating stimuli. In particular, widespread Kpna expression is present in and regulates OL differentiation, and OLP display differential changes in Kpna isotype expression from the onset of differentiation. In addition, the three major Kpna subtypes (P/α2, Q/α3, S/α1) show differential responses to different pro-myelinating factors (in our experiments, T3 and CNTF). Subtype P/α2, epitomized by Kpna2, decreases in expression following T3 and CNTF treatment, Q/α3 factors (Kpna4 and Kpna3) increase substantially in response to CNTF, and S/α1 members (Kpna1 and Kpna6) display responsiveness to T3 as well as to CNTF. Interestingly, while we have previously reported CNTF-mediated regulation of Kpna1 [22], our findings here identify additional sensitivity to T3 in primary cultures. Finally, loss of the highly CNTF-responsive Kpna4 results in inhibition of CNTF-induced differentiation, in the absence of changes in proliferation or viability. Collectively, these findings suggest that canonical nuclear import is an integral part of OL differentiation, and that specific Kpna paralogs may fulfill important functions downstream of pro-myelinating cues. The array of nuclear import proteins expressed in the OL lineage may relate to the presence of diverse cargo specificities [9] and to potentially different roles for the isotypes at various stages of development both during and outside of differentiation.
Kpnas have roles in many cellular processes [11,13,19,30,32]. The observation that four Kpna proteins increase their expression with differentiation while one decreases suggests that OL differentiation imposes a high demand for these import proteins and nuclear import more broadly. Indeed, successful OL differentiation requires a complex and coordinated intrinsic transcriptional program, and a diverse set of cargo is needed to enact this program [1]. In line with this, our data show that complete inactivation of canonical nuclear import strongly inhibits OL differentiation and, at higher levels of pharmacologic inhibition, results in programmed cell death, the latter also suggesting a potential role for nuclear import in maintenance of cell viability.
Interestingly, Kpnas may also play specific roles downstream of different extrinsic pro-myelinating cues. In our study, Kpna4 appears indispensible to OL differentiation induced by CNTF, but has a much more minor impact on T3-mediated differentiation. It is quite possible that it confers CNTF-specific roles, shuttling transcription factors that play a role in Jak-Stat signaling, the major signaling pathway downstream of CNTF [33]. Indeed, Kpna4 has been shown to carry Stat3 as a cargo in other cell types, even independent of its phosphorylated Kpna4 immunoreactivity (B, white arrowheads). (D) qPCR data from OLP nucleofected with siKpna4 or siNT control for 24h. Silencing is efficient and is selective for Kpna4. (E,F) Representative confocal image and associated quantification from primary mouse OLP silenced for Kpna4 versus nontargeting control, then allowed to proliferate in the presence of the mitogens PDGFAA and bFGF for 24h. Active proliferation was assessed by immunocytochemistry for Ki67, confocal imaging, and quantitation of %Ki67/DAPI cells. (G-I) To directly compare responses of Kpna4-deficient cells and controls to CNTF vs. T3 in parallel, primary cultures were nucleofected with siRNA for Kpna4 or nontargeting (NT) control, then treated with either CNTF or T3. (G) Representative confocal image and associated quantification from primary mouse OLP nucleofected with siKpna4 or siNT control and then differentiated with T3 (60 ng/ml) or CNTF (100 ng/ml) for 72h. This image shows maturation markers for OLP (NG2), immature/mature OL (O4), and mature OL (MBP) in the Olig2 + OL lineage in Kpna4-deficient cultures and controls treated with T3 or CNTF. Maturation was assessed by quantifying %O4/DAPI (H) and %MBP/DAPI (I) expressing cells. Maturation was reduced in CNTF-treated cultures. In T3-treated cultures, only a slight reduction was observed in the proportion of MBP expressing cells, which did not reach significance, illustrating a differential impact of silencing Kpna4 depending on the growth factor added to induce differentiation. (J,K) The reduction in maturation markers coincided with a loss in complexity of branching morphology, a marker of OL maturity, as assessed by fractal analysis.  Scalebars, (A) 50μm, 25μm inset, (B) 10μM, (C) 25μm, (E) 20μm, (G) 20μm. Individual data values are in S1 data. (activated) state [34], and Kpna proteins also carry Klf transcription factors [35,36], such as Klf6, which we have shown to be linked to Jak-Stat signaling in CNS myelination [24]. This suggests that at least one Kpna may have gp130 pathway-specific, or Jak-Stat-specific, functions. A wider precedent also exists for Kpna-dependency of Stat proteins, although not necessarily in the context of CNTF-gp130 signaling. For example, it has been shown that Kpna1 mediates type I interferon-induced ISGF3 (Stat1/Stat2/IRF9) complex nucleocytoplasmic import [37]. In addition, Kpna1 is also implicated in Stat1 import in response to type II interferon signaling activation [38]. Other paralogs may similarly function downstream of and in coordination with specific extrinsic cues, warranting further exploration. A more useful understanding of specific Kpna paralog functions will also derive from examination of their unique cargo affinities. Although outside the scope of the current work, future studies could combine co-immunoprecipitation of the different paralogs with mass spectrometry at various stages of the OL lineage, to obtain a more complete accounting of their cargo. Such work should not only implicate the Kpna isotypes in known and essential OL differentiation mechanisms, but may also identify, in addition to novel transcription factors, regulators, and epigenetic modifiers that can play important roles in development and CNS myelination. Such findings could also potentially be extended to studies in remyelination, to improve our understanding of the mechanisms that control myelin repair and return of function in demyelinating conditions such as MS.
Examination of Kpnas in the context of the OL lineage has been a relatively under-represented area, and we have begun to examine their broader roles in OL development, particularly in differentiation. While recent work from our laboratory clearly demonstrates a role for Kpna1 in OL differentiation and viability [24], the current study provides evidence for broader Kpna paralog involvement in OL differentiation. Studies of karyopherins may also inform development of new treatments for disease. Pharmacologic inhibitors or activators of specific isotypes, should they be developed, could have significant therapeutic value. For example, if future studies identify a specific Kpna or subtype as responsible for importing a differentiation inhibitor into the nucleus in the context of stalled remyelination in MS, a drug inhibiting its nucleocytoplasmic transport could produce beneficial therapeutic outcomes. Indeed in other contexts, such as cancer, specific nuclear import inhibitors are being studied as effective treatments [39][40][41][42]. Collectively, findings from this study implicate differential nuclear import as an important factor regulating OL differentiation. Elucidating the functions of this canonical mechanism in both physiology and disease has exciting implications for the generation of new treatments in white matter disorders.

Mice
Work was approved by Institutional Animal Care and Use Committee at the Icahn School of Medicine at Mount Sinai under protocol numbers LA13-00029 and LA12-00235 and adhered to the American Veterinary Medical Association guidelines. Our institution has an Animal Welfare Assurance on file with the Office for Laboratory Animal Welfare. The Assurance number is A3111-01. All mice used in this study were C57BL6 from Jackson Laboratories (Bar Harbor, Maine).

Inhibition of canonical import
The Kpnb inhibitor importazole was obtained from EMD Bioscience, and used at concentrations described in the text. Growth factors were added on top of inhibitors, maintaining described concentrations. OLP were pre-treated for 1h with concentrations of importazole indicated in the text or DMSO before the addition of T3.

Transfection
Immunoblotting Protein extracts were separated by sodium dodecyl sulfate-polacrylamide gel electrophoresis (SDS-PAGE) and transferred to onto PVDF membranes (Miliipore, Billerica, MA, USA) using a buffer continaing 25mM Tris base, pH 8.3, 192 mM glycine, 20% (vol/vol) methanol for 1h at 100 V at 4˚C. Membranes were blocked for 1h in 10% milk/0.1% Tween/TBS, then incubated overnight at 4˚C with primary antibodies as described in the text. After incubation in primary antibody, membranes were rinsed with 0.1% Tween/TBS three times, and subsequently incubated 2h at room temperature with secondary light-chain specific HRP-conjugated antibodies diluted 1/1000 in 10% milk/0.1% Tween/TBS. After rinsing, membrances were incubated with ECL (Thermo Scientific, Somerset, NJ) for 10 min and then revealed. For densitometry, unsaturated films were scanned using a Canon LiDE scanner (Lake Success, NY), with mean pixel density of each band measured using ImageJ software. Measurements were standardized to actin loading control, and fold change versus baseline calculated.

Immunofluorescence
Frozen 25μm sections of paraformaldehyde-perfused CNS were immunostained. Sections were soaked in 100˚C citrate buffer pH 6 for 15 min and then let to cool in the buffer to room temperature and washed three time in PBS. Samples were then incubated in blocking buffer (10% normal goat serum in PBS/Triton-X 0.3%) for 1h at room temperature, and then incubated with primary antibodies for markers as specified in the text, diluted in blocking buffer overnight at 4˚C. After rinsing three times with PBS/Triton-X 0.3%, samples were incubated with appropriate AlexaFluor conjugated secondary antibody in blocking buffer for 1h at room temperature. Following rinsing, cells were mounted using DAPI Fluoromount-G or Fluoromount-G without DAPI (SouthernBiotech, Birminhgam, AL). Samples were examined using a Zeiss LSM880 confocal microscope and stacks collected using 1-2μm on the Z-axis and assembled into projections. Cells were counted in at least four stacks per condition/per marker/per timepoint in each experiment by a blinded observer using ImageJ software.

Fractal analysis
Fractal analysis was conducted on cells expressing and immunostained for the O4 antigen which identifies immature and mature OL. Image processing and analysis was done in ImageJ software using the FracLac plugin with default settings. 15 cells per well were analyzed. For proper analysis, assessed cells were isolated without contact with other cells. Imaged cells were binarized by thresholding with default settings and then thinned using the "Skeletonize" command (a thinning algorithm within ImageJ). The box-counting fractal dimension (Db) was calculated for each cell using the FracLac plugin at default settings. This value is derived from the slope of the linear regression between the log of the number of boxes occupied by pixels and the log of these boxes' size.

RNA isolation and qPCR analysis
OLP lysate was prepared by the addition of 350 μl RLT solution (1ml RLT buffer, 10μl 2-Mercaptoethanol). Lysate was either used immediately or stored at -80˚C for later use. RNA was isolated from cultures using a modified RNeasy Mini kit (QIAGEN) protocol with homogenization via QIAshredder (QIAGEN) and the removal of genomic DNA with RNase-free DNase (QIAGEN). RNA quality was assessed using a Nanodrop spectrophotometer (Wilmington, DE, USA). RNA was reverse transcribed in 20 μL using SuperScript RT-PCR kit (Invitrogen) and subsequently diluted to 0.5 ng/μL for qPCR. qPCR was performed using PerfeCTa SYBR