Effect of Myelin Oligodendrocyte Glycoprotein (MOG35-55) on Cell Differentiation and Abzymes Production in Transgenic EAE-Prone Th Mice with T Cells Response During the Development of Experimental Encephalomyelitis

Background: The mechanisms of multiple sclerosis development are still unknown. It was shown that the development of experimental autoimmune encephalomyelitis (EAE) in EAE 29 prone C57BL/6 mice (model mimicking human multiple sclerosis) having B and T lymphocyte 30 responses is associated with modification in the differentiation profiles of bone marrow 31 hematopoietic stem cells (HSCs) and the increase in lymphocyte proliferation. Methods: Only T cell responses characterize other EAE transgenic prone Th mice. Different 33 characteristics of the autoimmune reaction in Th mice were analyzed. During the development of EAE (and inflammation processes), the differentiation profiles of Th mice bone marrow 35 HSCs (BFU-E, CFU-E, CFU-GM, CFU-GEMM, T, and B lymphocytes) were noticeably or 36 significantly different in male and female mice before and after their immunization with myelin 37 oligodendrocyte glycoprotein (MOG 35-55 ). 38 Results: The patterns of B and T (including CD4 and CD8 cells) lymphocytes proliferation 39 in several organs (spleen, thymus, bone marrow, blood, and lymph nodes) during 40 spontaneous (completely untreated mice) and MOG-treatment-accelerated development of 41 EAE was also remarkably or significantly different in male and female mice. All these 42 changes in male and female mice, despite some differences, were coupled with the increase 43 in the concentrations of autoantibodies against DNA, myelin basic protein, and MOG, and 44 with the increase in the relative activity of catalytic antibodies hydrolyzing these antigens. 45 Conclusions: A comparison of the changes in a large number of parameters characterizing 46 the development of EAE in Th and C57BL/6 mice was carried out. It was shown that MOG 47 very much accelerates the development of EAE in Th mice with T cell responses. Despite some 48 differences, the general patterns of the developing of spontaneous and MOG-accelerated 49 EAE in Th male and female mice and in C57BL/6 mice are similar to a notable extent. 50

there was analyzed the development of EAE using only male mice [7,8,34,35]. However, 138 multiple sclerosis is less common for men than for women [40]. Only about a third of 139 patients with MS are men, and the rest are women. In addition, the disease often occurs 140 quite late in men and proceeds more severely. 141 Therefore, taking this into account, in this study, we compared the development of 142 EAE after immunization with MOG of male and female Th mice. In addition, to compare 143 specific features of the EAE development in mice with T (Th) response and T and B 144 (C57BL/6) responses, some previously obtained data on the analysis of C57BL/6 male mice 145 were used [7,8,34,35]. 146 147

Weight and proteinuria of EAE mice 148
We have first analyzed over time changes in the relative weight of Th mice before and 149 after they immunization with MOG (Fig. 1a). Interestingly, the immunization of male 150 and female mice with MOG had minimal effect on their weight over time compared 151 with the untreated mice. After the immunization, a decrease in weight of C57BL/6 152 males during 10-15 days (Supplementary S1; all supplementary Figs are given in 153 Supplementary data) was more pronounced than in Th males (Fig. 1a). The overtime changes in proteinuria of Th mice before and after their immunization 163 with MOG were analyzed (Fig. 1b). At time zero (3 months of age) in two groups of seven 164 male and female mice before their immunization, the average urine protein concentration 165 was surprisingly very different: 7.6  2.2 and 3.2  1.0 mg/ml (P < 0.05), respectively (Fig.  166 1b). During the spontaneous development of EAE, the concentration of proteins in the 167 urine of Th males increased 1.6 times (to 12 mg/ml), while in females, only 1.3 times (to 168 4.2 mg/ml); the difference in concentration is statistically significant (P = 0.01). After 169 immunization of male mice, proteinuria was increased up to day 43, similar to its growth 170 before mice treatment, and then it began to decreased (Fig. 1b); the difference is 171 statistically significant (P = 0.01). In contrast to Th male mice, the immunization of 172 C57BL/6 male mice led to a relatively flowing 1.8-fold increase in proteinuria after 173 immunization compared with the spontaneous development of EAE ( Supplementary Fig.  174 S1). A completely different situation was observed in Th female mice having at zero time 175 lower proteinuria. In female mice from 15 to 30 days, proteinuria was sharply increased 176 3.2-fold (to 10.3 mg/ml) and then was decreased (Fig. 1b). Thus, the immunization with 177 MOG differently affects the proteinuria of female and male mice, which might be possibly 178 associated with some specific features of EAE development in Th female and male mice. 179 180

Hematopoietic progenitor colony formation 181
It was shown that the differentiation profile of bone marrow stem cells in CBA and BALB 182 mice not predisposed to AIDs does not remarkably change for at least 1-2 years [9][10][11]. It 183 was shown that spontaneous in-time development of EAE in C57BL/6 mice (change in the 184 analyzed parameters of EAE in time before immunization of mice) results in significant 185 changes in the differentiation profile of bone marrow stem cells [7,8,34,35]. Therefore, we 186 first performed carried out a parallel analysis of possible changes in the stem cell 187 differentiation profiles of three-month-old EAE prone Th mice (7 males and 7 females; the maximum number of mice, which it was possible to analyze in such parallel experiments). 189 During the 80 days of MOG-induced EAE development, a significant increase in the 190 relative number of BFU-E colonies occurred in the bone marrow of Th male and female 191 mice in the period from 15 to 40 days (Fig. 2a). In the beginning, there was 2.0-fold 192 difference (P < 0.05) between BFU-E colonies of male and female mice, but during the 193 entire analysis period of EAE development did not differ significantly (P = 0.19) except 194 1.8-fold difference (P < 0.05) at day 42. Simultaneously, during the spontaneous 195 development of EAE, the relative number of these colonies gradually decreases (Fig. 2a) S2). After 20 days, the relative number of these colonies increases in both immunized 202 and untreated C57BL/6 male mice. Th male mice before immunization do not demonstrate 203 an increase, but a decrease in the number of these colonies over time, and after 204 immunization, the increase in the number of BFU-E colonies begins earlier, from about 10-205 15 days of experiments (Fig. 2a). 206 A significant difference in the direction of changes in CFU-E colonies was observed 207 between female and male Th mice during spontaneous EAE development (Fig. 2b). The 208 number of CFU-E colonies in male mice increases during 10 days and then slowly 209 decreases; only slight changes in the relative number of CFU-E colonies in Th female mice 210 are observed up to approximately day 35, and then their number remarkably increases. 211 There was no, however, the statistical difference in these average values during the 212 development of EAE in male and female mice (P > 0.05). Immunization with MOG leads 213 to a significant increase in CFU-E colonies in Th males and females with earlier and 214 stronger growth of these colonies number in male mice (Fig. 2b). A statistically significant 215 2-fold difference (P < 0.05) in the number of colonies in males and females after their 216 immunization with MOG was observed in the period from 14 to 21 days (onset and acute 217 phase of EAE). Despite the significant difference in the profiles of changes in the relative 218 average number of CFU-E colonies during the entire experiment, the difference in the 219 colonies for female and male mice was insignificant, P > 0.05. 220 C57BL/6 males showed a slow increase in the number of these colonies 221 ( Supplementary Fig. S2). Nevertheless, in the case of males of both Th and C57BL/6 lines, 222 their immunization with MOG results in a sharp increase in the number of these colonies 223 by days 10-20 (Figs 2b and Supplementary Fig. S2). 224 The change in the average number of CFU-GM colonies over time during spontaneous 225 development of EAE in female and male Th mice is complex (Fig. 2c) and characterized 226 by a significant difference (P = 0.012). There is a decrease in the CFU-GM colonies at 10-227 20 days, but a notable increase by day 35 and a further perceptible decrease. Th males 228 showed an opposite change in the number of CFU-GM colonies before and after their 229 immunization with MOG demonstrating a maximal 2.7-3.0-fold (P < 0.05) difference from 230 30 to 40 days. While before treatment, the number of colonies sharply decreased by 30-40 231 days after immunization of male mice, their number decreased significantly from 10 and 50 232 days (Fig. 2c). No very significant changes were observed before immunization of Th 233 female mice, but after their immunization with MOG, a sharp increase in their number 234 occurred by 30 days (P = 0.015). In other words, there is an oppositely directed change in 235 the CFU-GM colonies in Th males and females after their immunization -in males, a sharp 236 decrease, and in females, a powerful increase in the number of colonies. In contrast to a 237 sharp increase before and a sharp decrease of CFU-GM colonies after Th males 238 immunization with MOG ( Fig. 2c) Thus, the changes observed in Th male and female mice during spontaneous 264 development of EAE are different for CFU-E, CFU-GEMM, and T cells but, to some 265 extent, similar for BFU-E, CFU-GM, and B cells (Fig. 2). After immunization with MOG, 266 some similarity in the change in the number of colonies is observed for BFU-E, CFU-E, 267 CFU-GEMM, and B-cells but is opposite for CFU-GM colonies and T cells (Fig. 2). Immunization of Th mice with MOG leads to an approximately comparable strong 279 increase in the number of B lymphocytes in the blood of males and female mice from 7 to 280 30 days, followed by a decrease in their number (Fig. 3a). At 7 days after immunization, 281 the relative number of B-cells in female is 1.3 times higher (P < 0.05) than in male mice. 282 The number of B lymphocytes growth in the thymus of female and male mice 283 proceeds smoothly and almost similar (Fig. 3b). The curves of changes in the number of B 284 lymphocytes in Th female and male mice after their immunization with MOG are very 285 complex. At first, from 7 to 20 days, an increase in the number of B cells is observed 286 (which is noticeably stronger in female mice) than a decrease (30-40 days) and again their 287 growth (Fig. 3b). A very keen increase in the number of B-lymphocytes (P < 0.05) was 288 observed in female mice at 22 days. 289 The complex patterns of changes in the number of B lymphocytes in the spleen and 290 lymph nodes of Th male and female mice were slightly different. During the first 7-16 291 days, there was in both male and female mice a decrease in the number of B cells in the 292 spleen, but an increase in lymph nodes, followed by a slight rise in spleen but a decrease in 293 lymph nodes (Figs 3c and 3d). Mice immunization with MOG dramatically changed these 294 patterns. In male mice spleens, there was a sharp decrease in B lymphocytes by day 22 (P 295 < 0.05), and then there was a remarkable increase in their relative content (Fig. 3c). In the 296 spleen of female mice, there was a constant but salutatory increase in B lymphocytes. The 297 patterns of reduction in B cells in the lymph nodes of male and female mice from 7 to 30 298 days after immunization mice with MOG were, to some extent, similar (Fig. 3d). After 299 immunization, there was a strong decrease of 2.3-2.4 times (P < 0.05) in the relative 300 number of B-lymphocytes in 15 days in female and 22 days in male mice (Fig. 3d). 301 Taken together, the curves of the changes in the average relative concentration of B 302 cells in various organs before and after immunization were different, but the characters of 303 these changes in every organ of Th male and female mice were, to some extent similar 304 (Figs 2 and 3). 305 306

The content of T cells in different organs of mice 307
The patterns of changes in the number of T lymphocytes in the bone marrow of Th male 308 and female mice before and after the immunization were directly opposite. We evaluated 309 overtime dependencies in the changes average relative number CD4 and CD8 lymphocytes 310 in various organs before and after mice immunization with MOG. Th male mice before 311 immunization during spontaneous EAE demonstrated during 20 days a slight decrease in 312 the relative number of total CD4 cells in the bone marrow with access to the plateau after 313 20 days (Fig. 4a). After immunization, a stronger decrease in the relative number of CD4 314 was observed with its very sharp increase on day 22 (P < 0.05). During the spontaneous 315 development of EAE in Th female mice, the relative number of CD4 cells began to 316 increase only after 40-50 days compared to zero time (Fig. 4a). After immunization of Th 317 female mice, a smooth but finally very strong decrease in average CD4 lymphocytes in the 318 bone marrow (P = 0.015) in comparison with spontaneous ЕAE was observed (Fig. 4a). 319 During the spontaneous development of EAE, male and female mice demonstrated 320 similar patterns of changes in CD4 cells in the blood with a sharp increase at 7-15 days 321 after the start of the experiment. Immunization of Th male and female mice resulted in a 322 strong decrease in the relative number of CD4 lymphocytes in their blood and complex 323 jump-like patterns of changes in these cells over time (Fig. 4b). The difference in the 324 number of CD4 cells before and after immunization in males was statistically significant (P 325 = 0.02) but not in females (P = 0.09). The maximum 4.1-fold difference (P < 0.05) in this 326 parameter in females was observed at 8 days. sharp increase in the number of CD4 cells in these organs of male and female mice from 7 345 to 23 days, and then there was a very strong decrease in their relative number (Fig. 4e). The 346 maximum 1.3-fold (P < 0.05) difference in the number of CD4 cells in female and male 347 mice after immunization was observed at 15 days. During the spontaneous development of 348 EAE, the patterns of changes in CD8 cells in the bone marrow of Th mice are similar (Fig.  349 5a) to overtime changes in CD4 lymphocytes (Fig. 4a). 350 In the first 7-15 days, there is a noticeable decrease in their average relative number 351 for male but an increase for female mice (Fig. 5a). Then the Th male mice demonstrated a 352 plateau, while the female mice show an increase in CD8 cells after 40 days (Fig. 5a). 353 Interestingly, the patterns of overtime changes in the number of CD8 lymphocytes in the 354 bone marrow of Th male and female mice (Fig. 5a) also have a similar character to those 355 for CD4 cells (Fig. 4a). The number of CD8 cells in female mice sharply decreases by day 356 20 and reaches a plateau; the strong difference in average data before and after mice 357 immunization during complete experiments is observed (P = 0.003). In the case of male 358 mice, there is a strong decrease in the number of these lymphocytes following a sharp 359 splash in their number on day 22 (Fig. 5a). At 22 days, the difference in the number of cells 360 of immunized female and male mice is 12.5 times (P = 0.007). 361 The curves characterizing changes in CD8 and CD4 lymphocytes in the blood before 362 immunization of Th male and female mice are, to some extent, similar to a larger or 363 smaller increase in their number in 7 days with the following decrease in their number 364 (Figs 4b and 5b). After immunization of female mice, there is a very slight increase in CD8 365 lymphocytes with a temporary plateau when male mice show a sharp increase in the 366 number of these cells up to day 22 with their subsequent decrease (Fig. 5b). A maximal 2-367 fold difference (P < 0.05) in the number of CD8 cells in immunized female and male mice 368 is observed 22 days after their treatment. 369 The characters of the in-time changes in CD8 cells in the thymus of Th male and 370 female mice before immunization is somewhat close to that for these lymphocytes changes 371 in the bone marrow (Fig. 5c) and CD4 cells in the thymus (Fig. 4c). The curve 372 corresponding to overtime changes in CD8 in the thymus of male mice (Fig. 5c) after their 373 immunization almost repeats that for the bone marrow (Fig 5a). The nature of the changes in CD8 lymphocytes in lymph nodes is very similar for 390 male and female mice, both before and after their immunization with MOG (Fig. 5e). 391 During the spontaneous development of EAE, male and female mice in overall show an 392 almost smooth increase in the number of these cells over time. After immunization of male 393 and female mice, there is a sharp increase in the relative number of CD8 lymphocytes to 394 day 22 with their subsequent decrease (Fig. 5e). 395 Thus, the characters of the changes in CD4 and CD8 cells in all organs of Th male and 396 female mice are very complex. The greatest differences in the number of CD4 and CD8 397 cells are observed in the bone marrow of male and female mice after their immunization 398 with MOG (Figs 4a and 5a). In addition, there was observed a significant difference in the 399 dependencies of overtime changes in CD8 cells in male and female mice after their 400 immunization with MOG (Fig. 5b).  Fig. S3). During 73 421 days of spontaneous development of EAE, the anti-DNA Abs concentration in Th mice 422 increased 5.3-and 3.8-fold (P < 0.05) for female and male mice, respectively (Fig. 6a). 423 The treatment of Th mice with MOG led to a significant and comparable decrease in anti-424 DNA Abs concentration in sera of male and female mice during the onset and acute phase 425 (7-23 days) of EAE (at 14 days 3.6-4.0-fold, P < 0.05) (Fig. 6a). However, then, during the 426 period of remission (> 30 days), the concentration of anti-DNA antibodies was slowly 427 increased; the difference in the concentration of anti-DNA Abs before and after 428 immunization of mice with MOG during the entire analysis period was statistically 429 significant, P < 0.05. 430 Interestingly, unlike Th mice, immunization with MOG of C57BL/6 male mice did 431 not lead not to the decrease but to an increase in the concentration of anti-DNA in the 432 blood sera (Supplementary Fig. S3). As in the case of anti-DNA auto-Abs, immunization of male and female mice with MOG 438 led to a sharp decrease in the concentration of antibodies against MBP starting the first 439 seven days after mice immunization (at 7 days 1.6-2.0-fold, P < 0.05) (Fig. 6b). Wherein, 440 after immunization, a statistically significant difference was observed between the anti-441 MBP IgGs concentrations in female and male mice (P = 0.02). 442 As shown earlier, the blood of C57BL/6 mice, even at 3 months of age, contains Abs 443 against MBP, and their concentrations increase during spontaneous but much faster after 444 male immunization with MOG ( Supplementary Fig. S3) [7,8,34,35]. During the 445 spontaneous development of EAE in Th male and female mice, auto-Abs against MOG 446 increased almost linearly from 0.02 to 0.032-0.036 A 450, P < 0.05 (Fig. 6c). It was 447 reasonable to expect that immunization of mice with external MOG will lead to a sharp 448 increase in the concentration of antibodies against this peptide. However, we obtained an 449 unexpected result. The immunization of mice with MOG peptide led to a very powerful 450 suppression of the formation of antibodies against this peptide (from 7 to 40 days) in both 451 male and female mice (Fig. 6c)  At time zero, DNA-hydrolyzing activity of IgGs in Th female mice was approximately 502 by a factor of 2.5 higher than that for male mice (P < 0.05) (Fig. 8a). 503 During spontaneous development (73 days) of EAE by Th mice, the DNase activity of 504 female mice Abs was increased 2.8-fold, while for male mice, IgGs activity was risen by a 505 factor of 11 (P < 0.05) (Fig. 8a). The immunization with MOG of male mice led to an 506 increase in this activity today 30 by a factor of 32, while for female mice 11-fold, P < 0.05 507 (Fig. 8a). The DNase activity of female antibodies at 83 days after immunization was 2 508 times higher than that from the blood of males (P < 0.05). Approximately the same 509 situation was found earlier for C57BL/6 male mice; immunization with MOG led to an 510 increase in activity 25-fold at about day 20 after immunization of mice (Supplementary 511 Fig. S4). It should be noted that immunization of C57BL/6 male mice leads to a strong 512 increase in the concentration of anti-DNA antibodies and to a parallel increase in their 513 activity in DNA hydrolysis. At the same time, at a sharp decrease in the concentration of 514 antibodies against DNA in Th mice (Fig. 6a), there is a very strong increase in their activity 515 in the splitting of DNA (Fig. 8a). 516 The relative activity of IgGs from Th male and female mice at zero time in the 517 hydrolysis of MBP was nearly the same (Fig. 8b). There was, however, a strong slowdown 518 in the growth of abzyme MBP-hydrolyzing activity in sera of males compared to females 519 during the spontaneous development of EAE (Fig. 8b). In male mice after immunization, 520 there was a sharp increase in the hydrolysis of MBP antibodies at 7 days. But then, after 20 521 days, there was a decrease in MBP-hydrolyzing activity in male and female mice (Fig. 8b). 522 Finally, at 73 days after immunization, the activity of antibodies from the serum of male 523 and female mice was 23-26 times lower than before their treatment with MOG after 524 spontaneous development of EAE (P < 0.05). 525 The profiles of changes in the activity of antibodies from the blood of female and male 526 mice in the hydrolysis of MOG before as well as after immunization of mice are very 527 similar (Fig. 8b). The increase over time in the relative concentrations of Abs against MBP 528 in male and female mice during spontaneous development of EAE correlates with the 529 increase in the activity of abzymes hydrolyzing MBP (Figs 6b and 8b). At the same time, a 530 keen decrease in the concentration of antibodies against MOG (Fig. 6c) after the 531 immunization was inversely related to the significant increase in the relative activity of 532 these antibodies in the hydrolysis of MOG (Fig. 8c). 533 Before and after immunization of C57BL/6 male mice, there were parallel increases in 534 the concentration of anti-MOG antibodies and their activities in MOG hydrolysis 535 ( Supplementary Figs S3 and S4). 536 Thus, some noticeable differences are apparent for male and female mice, not only in 537 the dependencies of changes over time in differentiation profiles of stem cells (Fig. 1) and 538 patterns of changes in lymphocyte proliferation in different organs (Figs 2-4), but also in 539 relative concentrations of Abs against DNA, MBP, and MOG, as well as IgG activities in 540 their hydrolysis (Figs 6 and 8). The activated myelin-reactive B and T cells are both important for MS pathogenesis in 549 mammals [1][2][3]. Therefore, it was interesting to analyze the changes of pathology 550 development important parameters during the spontaneous and MOG-accelerated 551 development of EAE in Th mice, demonstrating a T lymphocyte response [33]. 552 The main difference in the differentiation profiles of HSCs during spontaneous 553 development of EAE in Th male and female mice was revealed for CFU-E, CFU-GEMM, 554 and T cells. However, the differentiation of BFU-E, CFU-GM, and B cells was, to some 555 extent, similar (Fig. 2). After mice immunization with MOG, a certain similarity in the 556 differentiation was observed for BFU-E, CFU-E, CFU-GEMM, and B-cells, but it was 557 reverse directions for CFU-GM colonies and T cells (Fig. 2). Some differentiation profiles 558 of HSCs in male C57BL/6 mice ( Supplementary Fig. S2) in overall were different in 559 comparison with Th mice (Fig. 2). 560 After immunization of mice with MOG in nearly all organs of Th male and female 561 mice similar to C57BL/6 mice, was observed a sharp increase in the relative number of 562 lymphocytes in the period from 7 to 23 days corresponding to the onset and acute stages of 563 pathology (Figs 4 and 5). The relative content of B and T lymphocytes in various organs and blood of mice was 704 estimated by flow cytometry. Peripheral blood was obtained using mice's standard 705 decapitation. Sodium citrate was used as an anticoagulant. Five hundred thousand 706 leukocytes in not more than 150 μL were used for cytometric analysis. Cells were 707 incubated with monoclonal Abs for 20 min in the darkness, then the cells of blood samples 708 were lysed using a special 10-fold volume of RBC lysis buffer for 20 min, centrifuged for 709 10 min, and washed using 500 μL of PBS buffer (150 mM NaCl, 17 mM KH 2 PO 4 , 52 mM 710 Na 2 HPO 4 ) containing 0.02% EDTA and 1% sodium azide. After samples centrifugation, 50 711 μL of PBS buffer was added to the cell pellet, and they were analyzed by a flow cytometer. 712 Lymphocytes were isolated from blood, bone marrow, thymus, lymph nodes, and 713 spleen. Bone marrow was derived by rinsing the femoral cavity. Lymph nodes and thymus 714 samples were carefully homogenized, large particles were removed, and different cells 715 were resuspended by their passing using a disposable syringe through a needle. Spleen More than 100,000 events were collected for each sample. Gating was performed as 734 follows: the total cell population was isolated according to cell size and granularity, and the 735 white blood cell population was determined using the pan-white blood cell marker CD45+ 736 and populations of CD3+ and CD3-leukocytes were obtained. In the CD3+ leukocytes 737 population (T cells), CD4+ and CD8+T cells were determined, and in the CD3-leukocytes 738 population, the relative content of CD19+ B cells was found. For all groups, the percentage 739 ratio was calculated relative to the initial lymphocyte population. 740 741

Statistical analysis 742
The final determined values are given as the mean±S.D. of three independent experiments for each 743 mouse (7 male and 7 female mice); the data were averaged over seven different male and female 744 mice of each group. Some of the sets of samples did not match the Gaussian distribution. Therefore, 745 to estimate the differences between parameters analyzed, the Mann-Whitney U test was utilized; P < 746 0.05 was regarded statistically significant.     Overtime changes in CD4 lymphocyte proliferation (%) in bone marrow (A), blood (B), thymus (C), spleen (D), and lymph nodes (E) for untreated and MOG-treated mice. The error in CD4 lymphocyte determination from three independent experiments in the case of each mouse of every group (7 mice) did not exceed 7-10%. Overtime changes in CD8 lymphocyte proliferation (%) in bone marrow (A), blood (B), thymus (C), spleen (D), and lymph nodes (E) for untreated and MOG-treated mice. The error in CD4 lymphocyte determination from three independent experiments in the case of each mouse of every group (7 mice) did not exceed 7-10%.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarydata.pdf