Sex-specific IL-6-associated signaling activation in ozone-induced lung inflammation

Ozone-associated lung inflammation and up-regulation of IL-6 and IL-6R expression

With one-way analysis of variance, ozone exposure resulted in a significant increase
in the expression levels of IL-6 in both male and female mice compared to the matched
controls exposed to FA (Fig. 1a, b, Additional file 1: Figure S1-a). However, the expression of IL-6 was significantly higher in females
in comparison to males exposed to O
₃
(Fig. 1a, b). A two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 52)?=?6.55, p?=?0.013, partial ?2?=?0.112. An analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in IL6
expression score between filter air and ozone exposure. Ozone-exposed females had
a significant increase in the IL6 score of 0.566 (95 % CI, 0.313 to 0.819) points
compared to the ozone-exposed males, F (1, 52)?=?20.11, p?=?0.0005, partial ?2?=?0.279 (Table 1). Similarly, females with ozone exposure had significant increase in the IL6 score
of 2.641 (95 % CI, 2.38 to 2.89) points compared to the females exposed to the filter
air, F (1, 52)?=?438.0, p?=?0.0005, partial ?2?=?0.894 (Table 1). Interaction effect of sex and exposure for IL6 expression is given in Fig. 1c, d.

Fig. 1. IL6 and IL6R expression and effect of ozone exposure. Left panels: a Representative Western blot images of IL6 and IL6R expression in males and females
with filter air and O
₃
exposure; univariate analysis of IL6 (b) and IL6R (e), expression in males and females, with FA and O
₃
exposure; two-way ANOVA interaction effect of sex (c) and exposure (d), for IL6 expression and sex (f) and exposure (g), for IL6R expression. Right panels: h Representative Western blot images of IL6 and IL6R expression in estrous cycle stages
of females, with filter air and O
₃
exposure; univariate analysis of IL6 (i) and IL6R (l), expression in estrous cycle stages of females, with FA and O
₃
exposure. Two-way ANOVA interaction effect of exposure (j) and estrous cycle stages (k), for IL6 expression and exposure (m) and estrous cycle stages (n), for IL6R expression. Univariate analysis data expressed as Ranks-Kruskal-Wallis
test of densitometric analysis; the values are depicted as mean with SD, where *p???0.05, **p???0.01, and ***p???0.001 are the levels of statistical significance compared to controls (n?=?6–8 per group). Two-way ANOVA for IL6 and IL6R analysis is given in Tables 1 and 3, respectively

Table 1. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for IL6 expression

To evaluate the contributions of female hormones to this regulation, we further investigated
IL-6 levels in females exposed to O
₃
or FA at different stages of the estrous cycle. Serum hormone measurements performed
at the time of sample collection (between 6:00 pm and 7:00 pm) indicated higher levels
of luteinizing hormone and progesterone in proestrus vs. other estrous cycle stages
(Table 2), as previously described 47]. Thus, for the purpose of this work, we performed comparisons of lung protein expression
levels in proestrus vs. the rest of the days combined. With one-way analysis of variance,
we found that exposure to O
₃
on the day of proestrus had a slightly higher but non-significantly different effect
on lung IL-6 expression than exposure in non-proestrus cycle stages (Fig. 1h, i); however statistically, the O
₃
exposed non-proestrus cycle stage had a significant increase in the expression due
to much lower IL-6 levels in the matched FA-exposed group. A two-way ANOVA examining
the effects of estrous cycle stages and filter air/ozone exposure on the IL6 expression,
showed statistically significant interaction F (1, 24)?=?4.45, p?=?0.045, partial ?2?=?0.156. An analysis of simple main effects for estrous cycle
stages and exposure type with Bonferroni adjustment revealed statistically significant
difference in IL6 expression score. Proestrus females exposed to ozone had a significant
increase in the IL6 score of 2.462 (95 % CI, 1.97 to 2.95) points compared to the
filtered air-exposed proestrus females, F (1, 24)?=?108.67, p?=?0.0005, partial ?2?=?0.819 (Table 1). Similarly, ozone-exposed proestrus females had a significant increase in the IL6
score of 1.245 (95 % CI, 0.757 to 1.732) points compared to the non-proestrus females
exposed to ozone, F (1, 24)?=?27.77, p?=?0.0005, partial ?2?=?0.536 (Table 1). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the IL6 is given in Table 1. Interaction effect of exposure and estrous cycle stages for IL6 expression is given
in Fig. 1j, k.

Table 2. Serum hormone levels in female mice

Levels of IL-6R exerted an overall increase (Additional file 1: Figure S1-b) and a marked difference in expression between male and female mice
exposed to O
₃
or FA (Fig. 1a, e). With one-way analysis of variance, the basal levels of IL-6R in female mice exposed
to FA were lower compared to male mice. With O
₃
exposure, we found little to no change in the expression levels of IL-6R in male mice,
while a very significant increase in female mice (Fig. 1a, e). A two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 52)?=?37.94, p?=?0.0005, partial ?2?=?0.422. An analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in IL6R
expression score between filter air and ozone exposure. Ozone-exposed females had
a significant increase in the IL6R score of 1.409 (95 % CI, 0.984 to 1.835) points
compared to the ozone-exposed males, F (1, 52)?=?44.13, p?=?0.0005, partial ?2?=?0.459 (Table 3). Similarly, females with ozone exposure had a significant increase in the IL6R score
of 1.785 (95 % CI, 1.36 to 2.211) points compared to the females exposed to the filter
air, F (1, 52)?=?70.84, p?=?0.0005, partial ?2?=?0.577 (Table 3). Interaction effect of sex and exposure for IL6R expression is given in Fig. 1f, g.

Table 3. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for IL6R expression

In concordance with the IL-6 expression, one-way analysis of variance of female mice
exposed to O
₃
at different estrous cycle stages showed a slightly higher, but not significant, expression
of IL-6R in proestrus compared to non-proestrus (Fig. 1h, l). A two-way ANOVA examining the effects of estrous cycle stages and filter air/ozone
exposure on the IL6R expression, showed statistically insignificant interaction F (1, 24)?=?0.076, p?=?0.785, partial ?2?=?0.003. However, an analysis of simple main effects for estrous
cycle stages and exposure type with Bonferroni adjustment revealed statistically significant
difference in IL6R expression score. Non-proestrus female exposed to ozone had a significant
increase in the IL6R score of 2.592 (95 % CI, 2.104 to 2.954) points compared to the
filtered air-exposed non-proestrus female, F (1, 24)?=?150.94, p?=?0.0005, partial ?2?=?0.863 (Table 3). Likewise, filtered air-exposed proestrus females had a significant increase in
the IL6R score of 0.659 (95 % CI, 0.234 to 1.084) points compared to the non-proestrus
females exposed to the filter air, F (1, 24)?=?10.240, p?=?0.004, partial ?2?=?0.299 (Table 3). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the IL6R is given in Table 3. Interaction effect of exposure and estrous cycle stages for IL6R expression is given
in Fig. 1m, n.

Ozone-associated lung inflammation and expression of STAT3 unphosphorylated and STAT3
Serine 727 and Tyrosine 705 phosphorylation

In order to establish the possible role of STAT3 in relation to the IL-6 and IL-6R
expression, we further investigated the expression levels of STAT3 unphosphorylated,
as well as STAT3 Serine 727 (p-S727) and Tyrosine 705 (p-Y705) phosphorylation in
mice exposed to O
₃
or FA. The expression levels were further compared between males and females, and
the impact of exposure in different female estrous cycle stages was assessed.

One-way analysis of variance revealed no difference in the levels of lung unphosphorylated
STAT3 with O
₃
or FA exposure (Additional file 1: Figure S1-c). Similarly, no sex differences were observed in either groups (Fig. 2a, b). Similarly, a two-way ANOVA on the sex and ozone/filter air exposure showed statistically
insignificant interaction F (1, 52)?=?1.765, p?=?0.190, partial ?2?=?0.033. However, analysis of simple main effects for sex and
exposure type with Bonferroni adjustment revealed statistically significant but marginal
difference in unphosphorylated STAT3 expression score between filter air and ozone
exposure. Ozone-exposed females had a very marginal but significant increase in the
STAT3 score of 0.135 (95 % CI, 0.003 to 0.267) points compared to the ozone-exposed
males, F (1, 52)?=?4.22, p?=?0.045, partial ?2?=?0.075 (Table 4), whereas males with filter air exposure had an insignificant increase in the STAT3
score of 0.098 (95 % CI, 0.034 to 0.230) points compared to the males exposed to the
ozone, F (1, 52)?=?2.217, p?=?0.143, partial ?2?=?0.041 (Table 4). Interaction effect of sex and exposure for unphosphorylated STAT3 expression is
given in Fig. 2c, d.

Fig. 2. STAT3, STAT3 Serine 727, and STAT3 Tyrosine 705 phosphorylation and effect of ozone
exposure. Left panels: a Representative Western blot images of STAT3, STAT3 Serine 727 and STAT3 Tyrosine
705 phosphorylation expression in males and females with filter air and O
₃
exposure; univariate analysis of STAT3 (b), STAT3 Serine 727 (e), and STAT3 Tyrosine 705 (h), expression in males and females, with FA and O
₃
exposure; Two-way ANOVA interaction effect of sex (c) and exposure (d), for STAT3 expression, sex (f) and exposure (g), for STAT3 Serine 727 expression and sex (i) and exposure (j), for STAT3 Tyrosine 705 expression. Right panels: k Representative Western blot images of STAT3, STAT3 Serine 727, and STAT3 Tyrosine
705 phosphorylation expression in estrous cycle stages of females, with filter air
and O
₃
exposure; univariate analysis of STAT3 (l), STAT3 Serine 727 (o), and STAT3 Tyrosine 705 (r), expression in estrous cycle stages of females, with FA and O
₃
exposure. Two-way ANOVA interaction effect of exposure (m) and estrous cycle stages (n), for STAT3 expression, exposure (p) and estrous cycle stages (q), for STAT3 Serine 727 expression, and exposure (s) and estrous cycle stages (t), for STAT3 Tyrosine 705 expression. Univariate analysis data expressed as Ranks-Kruskal-Wallis
test of densitometric analysis; the values are depicted as mean with SD, where *p???0.05 and **p???0.01 are the levels of statistical significance compared to controls (n?=?6–8 per group). Two-way ANOVA for STAT3, STAT3 Serine 727, and STAT3 Tyrosine 705
phosphorylation expression analysis is given in Tables 4, 5, and 6, respectively

Table 4. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for STAT3 expression

However, one-way analysis of variance of different female estrous cycle stages exerted
a significant difference. We found that STAT3 unphosphorylated levels were significantly
decreased in females exposed to O
₃
in non-proestrus cycle stages, whereas females exposed in proestrus showed a marked
increase with O
₃
exposure compared to the matched control females (Fig. 2k, l). A two-way ANOVA examining the effects of estrous cycle stages and filter air/ozone
exposure on the unphosphorylated STAT3 expression showed statistically insignificant
interaction F (1, 24)?=?14.57, p?=?0.001, partial ?2?=?0.378. Analysis of simple main effects for estrous cycle stages
and exposure type with Bonferroni adjustment revealed statistically significant difference
in unphosphorylated STAT3 expression score. Proestrus females exposed to ozone had
a significant increase in the STAT3 score of 0.414 (95 % CI, 0.201 to 0.628) points
compared to the filtered air-exposed proestrus females, F (1, 24)?=?16.113, p?=?0.001, partial ?2?=?0.402 (Table 4). Likewise, ozone-exposed proestrus females had a significant increase in the STAT3
score of 0.588 (95 % CI, 0.375 to 0.801) points compared to the non-proestrus females
exposed to ozone, F (1, 24)?=?32.414, p?=?0.0005, partial ?2?=?0.575 (Table 4). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the unphosphorylated STAT3 is given in Table 4. Interaction effect of exposure and estrous cycle stages for unphosphorylated STAT3
expression is given in Fig. 2m, n.

When we compared he levels of STAT3 S727 phosphorylation in these mice, we found an
overall increase with O
₃
exposure (Additional file 1: Figure S1-d). These were comparable in males exposed to O
₃
or FA and in females exposed to FA. However, females exposed to O
₃
displayed a slight but not significant increase in the STAT3 p-S727 levels vs. females
exposed to FA (Fig. 2a, e). Two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 60)?=?13.568, p?=?0.0005, partial ?2?=?0.184. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in STAT3
S727 phosphorylation expression score between filter air and ozone exposure. Ozone-exposed
females had a significant increase in the STAT3 serine 727 score of 1.009 (95 % CI,
0.622 to 1.397) points compared to the ozone-exposed males, F (1, 60)?=?27.137, p?=?0.0005, partial ?2?=?0.311 (Table 5). Similarly, females with ozone exposure had significant increase in the STAT3 serine
727 score of 0.804 (95 % CI, 0.416 to 1.191) points compared to the females exposed
to the filter air, F (1, 60) =17.209, p?=?0.0005, partial ?2?=?0.223 (Table 5). Interaction effect of sex and exposure for STAT3 p-S727 expression is given in
Fig. 2f, g.

Table 5. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for STAT3 Serine 727 phosphorylation expression

With one-way analysis of variance, females exposed to O
₃
in proestrus depicted a markedly higher, but not significant, phosphorylation compared
to the females exposed in non-proestrus (Fig. 2k, o). Two-way ANOVA examining the effects of estrous cycle stages and filter air/ozone
exposure on the STAT3 serine 727 expression showed statistically significant interaction
F (1, 28)?=?38.710, p?=?0.0005, partial ?2?=?0.580. Simple main effects for estrous cycle stages and exposure
type with Bonferroni adjustment revealed statistically significant difference in STAT3
serine 727 expression score. Proestrus females exposed to ozone had a significant
increase in the STAT3 serine 727 score of 1.223 (95 % CI, 0.880 to 1.567) points compared
to the filtered air-exposed proestrus females, F (1, 28)?=?53.117, p?=?0.0005, partial ?2?=?0.655 (Table 5). Similarly, ozone-exposed proestrus females had a significant increase in the STAT3
serine 727 score of 1.995 (95 % CI, 1.652 to 2.339) points compared to the non-proestrus
females exposed to ozone, F (1, 28)?=?141.295, p?=?0.0005, partial ?2?=?0.835 (Table 5). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the unphosphorylated STAT3 serine 727 phosphorylation is given in Table 5. Interaction effect of exposure and estrous cycle stages for SATAT3 p-S727 expression
is given in Fig. 2p, q.

Univariate analysis of lung STAT3 Y705 phosphorylation revealed a significant overall
increase in animals exposed to O
₃
compared to FA (Additional file 1: Figure S1-e) irrespective of sex differences (Fig. 2a, h). Two-way ANOVA on the sex and ozone/filter air exposure also showed statistically
significant interaction F (1, 52)?=?12.40, p?=?0.001, partial ?2?=?0.193. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in STAT3
tyrosine 705 phosphorylation expression score between filter air and ozone exposure.
Ozone-exposed male had significant increase in the STAT3 tyrosine 705 score of 1.645
(95 % CI, 0.982 to 2.308) points compared to the ozone-exposed females, F (1, 52)?=?24.80, p?=?0.0005, partial ?2?=?0.323 (Table 6) whereas males with ozone exposure had insignificant increase in the STAT3 tyrosine
705 score of 4.214 (95 % CI, 3.35 to 4.87) points compared to the males exposed to
the filter air, F (1, 52)?=?162.686, p?=?0.0005, partial ?2?=?0.758 (Table 6). Interaction effect of sex and exposure for STAT3 p-Y705 expression is given in
Fig. 2i, j.

Table 6. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for STAT3 Tyrosine 705 phosphorylation expression

Ozone exposure in females in proestrus exerted a slightly higher expression change
in the STAT3 p-Y705 compared to females exposed in non-proestrus stages, but unilabiate
analysis exhibited no significant differences between these groups (Fig. 2k, r). Similarly, two-way ANOVA examining the effects of estrous cycle stages and filter
air/ozone exposure on the STAT3 tyrosine 705 expression showed statistically insignificant
interaction F (1, 28)?=?0.143, p?=?0.709, partial ?2?=?0.005. However, simple main effects for estrous cycle stages
and exposure type with Bonferroni adjustment revealed statistically significant difference
in STAT3 tyrosine 705 expression score. Non-proestrus female exposed to ozone had
significant increase in the STAT3 tyrosine 705 score of 1.185 (95 % CI, 0.866 to 1.504)
points compared to the filtered air-exposed non-proestrus female, F (1, 28)?=?58.045, p?=?0.0005, partial ?2?=?0.675 (Table 6) whereas filtered air-exposed proestrus females had significant increase in the STAT3
tyrosine 705 score of 0.617 (95 % CI, 0.299 to 0.936) points compared to the non-proestrus
females exposed to the filter air, F (1, 28) =15.759, p?=?0.0005, partial ?2?=?0.360 (Table 6). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the unphosphorylated STAT3 tyrosine 705 phosphorylation is given in Table 6. Interaction effect of exposure and estrous cycle stages for STAT3 p-Y705 expression
is given in Fig. 2s, t.

Ozone-associated lung inflammation and expression of JAK2 and JAK2 phosphorylation

Interleukin-6 preferentially activates STAT3 with phosphorylation of Y705 via the
JAK signaling pathway. To assess activation of this mechanism in our model, we measured
the levels of JAK2 and JAK2 phosphorylation (Y1007+Y1008) with relation to O
₃
and FA exposure and further comparison of sex differences and female estrous cycle.

Irrespective of sex differences, one-way analysis of variance of pooled data showed
a significant increase in the expression of JAK2 unphosphorylated with O
₃
exposure compared to animals exposed to FA (Additional file 1: Figure S1-f). However, comparison of males and females represented deviations in
the expression patterns, where O
₃
exposure resulted in an overall decrease in JAK2 expression in males, while females
had a significant increase in JAK2 expression (Fig. 3a, b). Two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 52)?=?627.9, p?=?0.0005, partial ?2?=?0.924. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in unphosphorylated
JAK2 expression score between filter air and ozone exposure. Ozone-exposed female
had significant increase in the unphosphorylated JAK2 expression score of 2.395 (95 %
CI, 2.225 to 2.543) points compared to the ozone-exposed males, F (1, 52)?=?1190.013, p?=?0.0005, partial ?2?=?0.958 (Table 7) whereas females with ozone exposure had significant increase in the unphosphorylated
JAK2 score of 2.165 (95 % CI, 2.026 to 2.304) points compared to the females exposed
to the filter air, F (1, 52) =972.74, p?=?0.0005, partial ?2?=?0.942 (Table 7). Interaction effect of sex and exposure for JAK2 expression is given in Fig. 3c, d.

Fig. 3. JAK2 and JAK2 phosphorylated (Y1007+Y1008) expression and effect of ozone exposure.
Left panels: a Representative Western blot images of JAK2 and JAK2 phosphorylated (Y1007+Y1008)
expression in males and females with filter air and O
₃
exposure; univariate analysis of JAK2 (b) and JAK2 (e) phosphorylated (Y1007+Y1008), expression in males and females, with FA and O
₃
exposure; two-way ANOVA interaction effect of sex (c) and exposure (d), for JAK2 expression and sex (f) and exposure (g), for JAK2 phosphorylated (Y1007+Y1008) expression. Right panels: h Representative Western blot images of JAK2 and JAK2 phosphorylated (Y1007+Y1008)
expression in estrous cycle stages of females, with filter air and O
₃
exposure; univariate analysis of i JAK2 and l JAK2 phosphorylated (Y1007+Y1008), expression in estrous cycle stages of females,
with FA and O
₃
exposure. Two-way ANOVA interaction effect of exposure (j) and estrous cycle stages (k), for JAK2 expression and exposure (m) and estrous cycle stages (n) and for JAK2 phosphorylated (Y1007+Y1008) expression. Univariate analysis data expressed
as Ranks-Kruskal-Wallis test of densitometric analysis; the values are depicted as
mean with SD, where *p???0.05 and **p???0.01 are the levels of statistical significance compared to controls (n?=?6–8 per group). Two-way ANOVA for JAK2 and JAK2 phosphorylated (Y1007+Y1008) analysis
is given in Tables 7 and 8, respectively

Table 7. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for JAK2 expression

Univariate analysis of females at different estrous cycle stages revealed a higher
and significant increase in the levels of unphosphorylated JAK2 in females exposed
to O
₃
in proestrus compared to females exposed in non-proestrus stages and females exposed
to FA (Fig. 3h, i). Two-way ANOVA examining the effects of estrous cycle stages and filter air/ozone
exposure showed statistically significant interaction F (1, 36)?=?19.596, p?=?0.0005, partial ?2?=?0.352. Simple main effects for estrous cycle stages and exposure
type with Bonferroni adjustment revealed statistically significant difference in unphosphorylated
JAK2 expression score. Ozone-exposed non-proestrus females had significant increase
in the unphosphorylated JAK2 score of 1.795 (95 % CI, 1.603 to 1.987) points compared
to the non-proestrus females exposed to the filter air, F (1, 36)?=?360.191, p?=?0.0005, partial ?2?=?0.909 (Table 7). Similarly, proestrus female exposed to ozone had significant increase in the unphosphorylated
JAK2 score of 0.555 (95 % CI, 0.363 to 0.747) points compared to the ozone-exposed
non-proestrus female, F (1, 36)?=?34.392, p?=?0.0005, partial ?2?=?0.489 (Table 7). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the unphosphorylated JAK2 is given in Table 7. Interaction effect of exposure and estrous cycle stages for JAK2 expression is given
in Fig. 3j, k.

JAK2 phosphorylation (Y1007+Y1008) showed an overall significant increase with O
₃
exposure compared to FA (Additional file 1: Figure S1g). Females showed a higher and significant expression of phosphorylated
JAK2 (Y1007+Y1008) compared to O
₃
-exposed males (Fig. 3a, e). Two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 52)?=?23.991, p?=?0.0005, partial ?2?=?0.316. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in phosphorylated
JAK2 expression score between filter air and ozone exposure. Ozone-exposed female
had significant increase in the phosphorylated JAK2 expression score of 1.122 (95 %
CI, 0.779 to 1.465) points compared to the ozone-exposed males, F (1, 52)?=?43.108, p?=?0.0005, partial ?2?=?0.453 (Table 8) whereas females with ozone exposure had significant increase in the phosphorylated
JAK2 score of 1.538 (95 % CI, 1.195 to 1.881) points compared to the females exposed
to the filter air, F (1, 52)?=?80.967, p?=?0.0005, partial ?2?=?0.609 (Table 8). Interaction effect of sex and exposure for JAK2 p-(Y1007+Y1008) expression is given
in Fig. 3f, g.

Table 8. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for JAK2 phosphorylation (Y1007+Y1008) expression

However, as opposed to unphosphorylated JAK2, phosphorylated JAK2 (Y1007+Y1008) was
found to have a higher and significant increase in females exposed to O
₃
in non-proestrus stages vs. females exposed in proestrus (Fig. 3h, l). Two-way ANOVA examining the effects of estrous cycle stages and filter air/ozone
exposure showed statistically significant interaction F (1, 24)?=?93.334, p?=?0.0005, partial ?2?=?0.795. Simple main effects for estrous cycle stages and exposure
type with Bonferroni adjustment revealed statistically significant difference in phosphorylated
JAK2 expression score. Ozone-exposed non-proestrus females had significant increase
in the phosphorylated JAK2 score of 1.965 (95 % CI, 1.710 to 2.220) points compared
to the non-proestrus females exposed to the filter air, F (1, 24)?=?253.484, p?=?0.0005, partial ?2?=?0.914 (Table 8). Similarly, non-proestrus female exposed to ozone had significant increase in the
phosphorylated JAK2 score of 1.578 (95 % CI, 1.324 to 1.883) points compared to the
ozone-exposed proestrus female, F (1, 24)?=?163.543, p?=?0.0005, partial ?2?=?0.872 (Table 8). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the phosphorylated JAK2 is given in Table 8. Interaction effect of exposure and estrous cycle stages for JAK2 p-(Y1007+Y1008)
expression is given in Fig. 3m, n.

Ozone-associated lung inflammation and expression of NF-?B (p105/p50)

NF-?B activation is widely implicated in inflammatory conditions and is also known
to possess cross-talk with pathways that may influence IL-6 expression. IL-6 via STAT3
transcription effector mediates local vascular macrophage activation in lungs and
protection from oxidative stress. In addition, the NF-?B–IL-6 signaling pathway plays
multiple roles in initiating and sustaining vascular inflammation. Assessment of NF-?B
expression in our model showed sex differences in the lungs of mice exposed to O
₃
or FA. Ozone-exposed male mice showed a reduction in NF-?B expression, whereas females
exposed to O
₃
had a significant increase in the NF-?B expression (Fig. 4a, b). However, due to sheer increase in the expression of NF-?B in females, pooling male
and female data together masked the decrease in the expression in males, and an overall
increase in expression was found with O
₃
exposure (Additional file 1: Figure S1-h). Two-way ANOVA on the sex and ozone/filter air exposure showed statistically
significant interaction F (1, 52)?=?266.435, p?=?0.0005, partial ?2?=?0.837. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in NF-?B
(p105/p50) expression score between filter air and ozone exposure. Ozone-exposed female
had significant increase in the NF-?B (p105/p50) expression score of 0.970 (95 % CI,
0.855 to 1.085) points compared to the ozone-exposed males, F (1, 52)?=?287.977, p?=?0.0005, partial ?2?=?0.847 (Table 9), whereas females with ozone exposure had significant increase in the NF-?B (p105/p50)
score of 1.056 (95 % CI, 0.941 to 1.170) points compared to the females exposed to
the filter air, F (1, 52)?=?341.072, p?=?0.0005, partial ?2?=?0.868 (Table 9). Interaction effect of sex and exposure for NF-?B expression is given in Fig. 4c, d.

Fig. 4. NF-?B (p105/p50) expression and effect of ozone exposure. Left panel: a Representative Western blot images of NF-?B (p105/p50) expression in males and females
with filter air and O
₃
exposure; b univariate analysis of NF-?B (p105/p50) expression in males and females with FA and
O
₃
exposure; two-way ANOVA interaction effect of sex (c) and exposure (d) for NF-?B (p105/p50) expression. Right panel: e Representative Western blot images of NF-?B (p105/p50) expression in estrous cycle
stages of females, with filter air and O
₃
exposure; f univariate analysis of NF-?B (p105/p50) expression in estrous cycle stages of females
with FA and O
₃
exposure. Two-way ANOVA interaction effect of exposure (g) and estrous cycle stages (h) for NF-?B (p105/p50) expression. Univariate analysis data expressed as Ranks-Kruskal-Wallis
test of densitometric analysis; the values are depicted as mean with SD, where **p???0.01 is the level of statistical significance compared to controls (n?=?6–8 per group). Two-way ANOVA for NF-?B (p105/p50) expression analysis is given
in Table 9

Table 9. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages, and exposure for NF-?B (p105/p50) expression

Univariate analysis of females exposed in different estrous cycle stages showed no
overall difference in the expression patterns of the NF-?B with O
₃
exposure, compared to the matched controls exposed to the FA (Fig. 4e, f). Examining the effects of estrous cycle stages and filter air/ozone exposure through
two-way ANOVA also showed statistically insignificant interaction F (1, 32)?=?3.774, p?=?0.061, partial ?2?=?0.105. Simple main effects for estrous cycle stages and exposure
type with Bonferroni adjustment revealed statistically significant difference in NF-?B
expression score. Ozone-exposed non-proestrus females had significant increase in
the NF-?B score of 1.129 (95 % CI, 0.940 to 1.318) points compared to the non-proestrus
females exposed to the filter air, F (1, 32)?=?148.459, p?=?0.0005, partial ?2?=?0.823 (Table 9). Similarly, proestrus female exposed to filter had insignificant increase in the
NF-?B score of 0.175 (95 % CI, 0.013 to 0.363) points compared to the filtered air-exposed
non-proestrus female, F (1, 32)?=?3.581, p?=?0.068, partial ?2?=?0.101 (Table 9). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the NF-?B is given in Table 9. Interaction effect of exposure and estrous cycle stages for NF-?B expression is
given in Fig. 4g, h.

Ozone-associated lung inflammation and expression of AKT1

Studies have documented that IL-6/STAT3 signaling can regulate AKT1 activation and
that both JAK2 and AKT1 may play role in the activation of NF-?B 48], 49]. Initially believed to operate as components of distinct signaling pathways, several
studies have demonstrated that the NF-?B and AKT1 signaling pathways can converge
and play a crucial role in stress responses and inflammation 50]. In our experimental model, ozone exposure in pooled male and female mice resulted
in no differences in AKT1 expression compared to animals exposed to FA (Additional
file 1: Figure S1i). However, analysis of sex differences in lung AKT1 levels in response
to O
₃
exposure depicted a marked decrease in males compared to matched controls exposed
to FA, while females exposed to O
₃
showed a significant increase vs. FA (Fig. 5a, b).

Fig. 5. AKT1 expression and effect of ozone exposure. Left panel: a Representative Western blot images of AKT1 expression in males and females with filter
air and O
₃
exposure; b univariate analysis of AKT1 expression in males and females with FA and O
₃
exposure; two-way ANOVA interaction effect of sex (c) and exposure (d) for AKT1 expression. Right panels: e Representative Western blot images of AKT1 expression in estrous cycle stages of
females, with filter air and O
₃
exposure; f univariate analysis of AKT1 expression in estrous cycle stages of females with FA
and O
₃
exposure. Two-way ANOVA interaction effect of exposure (g) and estrous cycle stages (h) for AKT1 expression. Univariate analysis data expressed as Ranks-Kruskal-Wallis
test of densitometric analysis; the values are depicted as mean with SD where **p???0.01 and ***p???0.001 are the levels of statistical significance compared to controls (n?=?6–8 per group). Two-way ANOVA for AKT1 expression analysis is given in Table 10

Two-way ANOVA on the sex and ozone/filter air exposure showed statistically significant
interaction F (1, 52)?=?60.730, p?=?0.0005, partial ?2?=?0.539. Analysis of simple main effects for sex and exposure
type with Bonferroni adjustment revealed statistically significant difference in AKT1
expression score between filter air and ozone exposure. Ozone-exposed female had significant
increase in the AKT1 expression score of 0.919 (95 % CI, 0.751 to 1.086) points compared
to the ozone-exposed males, F (1, 52)?=?121.354, p?=?0.0005, partial ?2?=?0.70 (Table 10) whereas females with ozone exposure had significant increase in the AKT1 score of
0.670 (95 % CI, 0.503 to 0.837) points compared to the females exposed to the filter
air, F (1, 52)?=?64.526, p?=?0.0005, partial ?2?=?0.554 (Table 10). Interaction effect of sex and exposure for AKT1 expression is given in Fig. 5c, d.

Table 10. Two-way ANOVA for univariate and pairwise comparisons between gender, estrous cycle
stages and exposure for AKT1 expression

Examination of lung AKT1 levels in females exposed to O
₃
at different stages of the estrous cycle showed an increased expression in animals
exposed in both proestrus and non-proestrus, but only females exposed in proestrus
had a significant increase in AKT1 (Fig. 5e, f). The effects of estrous cycle stages and filter air/ozone exposure through two-way
ANOVA also showed statistically significant interaction F (1, 24)?=?42.745, p?=??0.0005, partial ?2?=?0.639. Simple main effects for estrous cycle stages and
exposure type with Bonferroni adjustment revealed statistically significant difference
in AKT1 expression score. Ozone-exposed proestrus females had significant increase
in the AKT1 score of 1.892 (95 % CI, 1.719 to 2.065) points compared to the proestrus
females exposed to the filter air, F (1, 24)?=?511.045, p?=?0.0005, partial ?2?=?0.955 (Table 10). Similarly, proestrus female exposed to ozone had significant increase in the AKT1
score of 0.620 (95 % CI, 0.448 to 0.793) points compared to the ozone-exposed non-proestrus
female, F (1, 24)?=?54.923, p?=?0.0005, partial ?2?=?0.696 (Table 10). Filter air/ozone exposure on each sex alone and estrous cycle type alone in relation
to the AKT1 is given in Table 10. Interaction effect of exposure and estrous cycle stages for AKT1 expression is given
in Fig. 5g, h.