β-Galactosylceramidase Deficiency Causes Upregulation of Long Pentraxin-3 in the Central Nervous System of Krabbe Patients and Twitcher Mice

Globoid cell leukodystrophy (GLD), or Krabbe disease, is a neurodegenerative sphingolipidosis caused by genetic deficiency of lysosomal β-galactosylceramidase (GALC), characterized by neuroinflammation and demyelination of the central (CNS) and peripheral nervous system. The acute phase protein long pentraxin-3 (PTX3) is a soluble pattern recognition receptor and a regulator of innate immunity. Growing evidence points to the involvement of PTX3 in neurodegeneration. However, the expression and role of PTX3 in the neurodegenerative/neuroinflammatory processes that characterize GLD remain unexplored. Here, immunohistochemical analysis of brain samples from Krabbe patients showed that macrophages and globoid cells are intensely immunoreactive for PTX3. Accordingly, Ptx3 expression increases throughout the course of the disease in the cerebrum, cerebellum, and spinal cord of GALC-deficient twitcher (Galctwi/twi) mice, an authentic animal model of GLD. This was paralleled by the upregulation of proinflammatory genes and M1-polarized macrophage/microglia markers and of the levels of PTX3 protein in CNS and plasma of twitcher animals. Crossing of Galctwi/twi mice with transgenic PTX3 overexpressing animals (hPTX3 mice) demonstrated that constitutive PTX3 overexpression reduced the severity of clinical signs and the upregulation of proinflammatory genes in the spinal cord of P35 hPTX3/Galctwi/twi mice when compared to Galctwi/twi littermates, leading to a limited increase of their life span. However, this occurred in the absence of a significant impact on the histopathological findings and on the accumulation of the neurotoxic metabolite psychosine when evaluated at this late time point of the disease. In conclusion, our results provide the first evidence that PTX3 is produced in the CNS of GALC-deficient Krabbe patients and twitcher mice. PTX3 may exert a protective role by reducing the neuroinflammatory response that occurs in the spinal cord of GALC-deficient animals.


Introduction
Globoid cell leukodystrophy (GLD), or Krabbe disease, is an autosomal recessive neurodegenerative sphingolipidosis caused by genetic deficiency of the lysosomal hydrolase β-galactosylceramidase (GALC) [1]. GALC degrades galactosylceramide (a major component of myelin) and other terminal β-galactose-containing sphingolipids, including β-galactosylsphingosine (psychosine). The pathogenesis of GLD has been proposed to arise from the accumulation of the neurotoxic metabolite psychosine present at high levels in the CNS of Krabbe patients [2][3][4][5]. The disease is characterized by neuroinflammation and loss of oligodendroglia and Schwann cells, leading to demyelination of the brain, spinal cord, nerve roots and peripheral nerves [6]. A pathognomonic feature of GLD is the presence Int. J. Mol. Sci. 2022, 23, 9436 2 of 17 in the white matter of globoid cells [7,8], giant multinucleated cells likely originated from resident microglia [9]. Clinically, GLD may manifest in early infancy with fatal neurological dysfunctions [6,10,11]. The standard of care for this disease is hematopoietic stem cell transplantation [12,13].
The acute phase protein long pentraxin-3 (PTX3) is an important regulator of peripheral innate immunity and a key mediator of inflammation during cardiovascular and cerebrovascular diseases [14]. Growing evidence points to the involvement of PTX3 in neurodegenerative disorders [15]. PTX3 expression is upregulated in the central nervous system (CNS) following pro-inflammatory cytokine stimulation [16], seizure-induced neurodegeneration [17], and ischemia [18]. Accordingly, peripheral PTX3 levels are increased after experimental stroke in mice [19], and plasma PTX3 levels correlate with mortality after ischemic stroke in humans [20]. In addition, serum levels of PTX3 have been proposed as a novel biochemical marker in Parkinson's disease [21]. Despite this evidence, no data are available about the expression and role of PTX3 in sphingolipid disorders, including GLD.
Here, PTX3 production was investigated in the brain of Krabbe patients and in the CNS of GALC-deficient twitcher (Galc twi/twi ) mice, an authentic murine model of GLD [22,23]. Analysis of brain samples from Krabbe patients demonstrated that macrophages and globoid cells are intensely immunoreactive for PTX3. Accordingly, twitcher mice are characterized by the progressive upregulation of Ptx3 expression along the CNS caudal-rostral axis. This was paralleled by the upregulation of various proinflammatory genes and M1-polarized macrophage/microglia markers, with no changes in the peripheral organs. Crossing of Galc twi/twi mice with transgenic PTX3 overexpressing animals (hPTX3 mice) caused an amelioration of clinical features and attenuated spinal cord inflammation in hPTX3/Galc twi/twi offspring when compared to Galc twi/twi littermates at postnatal day P35, when animals show clear neurologic defects. However, PTX3 overexpression did ameliorate the histopathological findings and did not affect psychosine accumulation in the CNS of GALC-deficient animals when evaluated at this late stage of the disease.
Our results demonstrate that PTX3 is produced in the CNS of GALC-deficient Krabbe patients and twitcher mice and point to a possible protective role of this immune modulator by reducing, at least in part, the neuroinflammatory response that characterizes GLD.

PTX3 Immunoreactivity in the Brain of Krabbe Patients
To assess whether GALC deficiency may cause PTX3 upregulation in human CNS, we analyzed autopsy brain cortex samples from nine Krabbe patients (Table S1) and two matched controls. Most of the Krabbe samples showed the classical histopathological hallmarks of the disease, such as diffuse demyelination along with marked loss of oligodendrocytes, intense reactive gliosis evidenced by the diffuse immunoreactivity of the astrocytic marker GFAP, and infiltration of globoid cells ( Figure 1A). Previous observations had shown that activated astrocytes and microglia represent a major source of PTX3 in the brain [16,24]. Accordingly, single immunostains indicated that GLD macrophages and globoid cells are diffusely and intensely immunoreactive for PTX3. Double immunostains confirmed that PTX3 is mainly expressed within macrophages and globoid cells in Krabbe samples, its positivity being related to the extent of pathology (i.e., presence of perivascular globoid cells) and of GFAP-positive astrocytic gliosis (Table 1), while GFAP/PTX3 double positive astrocytes were barely detected. Control brain samples did not show double positive PTX3 immunoreactive cells ( Figure 1B). Overall, these findings demonstrate that PTX3 is upregulated in the brain of GLD patients, its immunoreactivity being mainly expressed in monocyte-derived cells and related to the extent of the pathology. positive PTX3 immunoreactive cells ( Figure 1B). Overall, these findings demonstrate that PTX3 is upregulated in the brain of GLD patients, its immunoreactivity being mainly expressed in monocyte-derived cells and related to the extent of the pathology. Figure 1. PTX3 immunolocalization in the brain of Krabbe patients. (A) Representative images from Krabbe brain cortex specimens showing the classical histopathological hallmarks of the disease in the white matter, with diffuse infiltration of macrophages (globoid cells) mainly distributed around the blood vessels (left image, H&E). Immunostains for PTX3 (brown) and GFAP (brown, middle and right images, respectively) show intense and diffuse immunoreactivity of PTX3 and gliosis. (B) Double immunostainings from Krabbe and control brain cortex samples. The large majority of CD68 + monocyte-derived cells (blue) are intensely immunoreactive for PTX3 (brown) within the white matter of Krabbe patients (left image), while reactive GFAP + astrocytes (blue) expressing PTX3 (brown) are barely detected (middle left image; asterisk). Conversely, no PTX3 + or CD68 + cells were detected in these samples, except for rare circulating CD68 -PTX3 + monocytes within the vessels (middle right image), and PTX3 was not detected in resident GFAP + astrocytes from control matched brains (right image). Original magnification: 40× (A); 60× (B). Scale bars: 50 μm (A); 30 μm (B).

Ptx3 Upregulation in Twitcher CNS
Based on the data obtained in Krabbe patients, the expression of Ptx3 was investigated in the CNS of GALC-deficient twitcher (Galc twi/twi ) mice, an authentic murine model of GLD [22,23]. In a first set of experiments, Ptx3 expression was assessed by qPCR analysis in the cerebrum of littermate Galc wt and homozygous Galc twi/twi mice. Measurements were performed at postnatal day P15 before the onset of evident neurologic signs, at P24 when pathologic alterations were occasionally detectable, and at P35 when Galc twi/twi mice showed clear neurologic defects, including tremor and hind limb paralysis [23,25,26]. As shown in Figure 2A, the brain levels of Ptx3 transcripts were already increased at P15, to further increase at P24 and P35. This was paralleled by the upregulation of the inflammatory cytokines Tnfα, Cxcl1 and Il1α and of the CCAAT/enhancer binding protein delta (Cebpd) gene, which encodes for a transcription factor mediating PTX3 expression in astrocytes during neuroinflammation [27]. Accordingly, a progressive increase in the levels of the astrocytic marker of gliosis Gfap and the microglial marker ionized calcium-binding adaptor molecule-1 (Iba1) [28] occurred in Galc twi/twi cerebrum when compared to controls.

Ptx3 Upregulation in Twitcher CNS
Based on the data obtained in Krabbe patients, the expression of Ptx3 was investigated in the CNS of GALC-deficient twitcher (Galc twi/twi ) mice, an authentic murine model of GLD [22,23]. In a first set of experiments, Ptx3 expression was assessed by qPCR analysis in the cerebrum of littermate Galc wt and homozygous Galc twi/twi mice. Measurements were performed at postnatal day P15 before the onset of evident neurologic signs, at P24 when pathologic alterations were occasionally detectable, and at P35 when Galc twi/twi mice showed clear neurologic defects, including tremor and hind limb paralysis [23,25,26]. As shown in Figure 2A, the brain levels of Ptx3 transcripts were already increased at P15, to further increase at P24 and P35. This was paralleled by the upregulation of the inflammatory cytokines Tnfα, Cxcl1 and Il1α and of the CCAAT/enhancer binding protein delta (Cebpd) gene, which encodes for a transcription factor mediating PTX3 expression in astrocytes during neuroinflammation [27]. Accordingly, a progressive increase in the levels of the astrocytic marker of gliosis Gfap and the microglial marker ionized calcium-binding adaptor molecule-1 (Iba1) [28] occurred in Galc twi/twi cerebrum when compared to controls. In keeping with the neuroinflammatory response that occurs in twitcher brain, we observed an increase in the levels of expression of the classical M1 macrophage polarization/microglia markers Fc receptor, IgG, low affinity III and II (Fcgr3/CD16 and Fcgr2/CD32) [29] with no changes in the expression of the alternative M2 macrophage polarization markers arginase-1 (Arg1) and mannose receptor C-type1 (Mrc1/CD206) [30]. A similar transcriptional profile was observed in the cerebellum ( Figure S1) and spinal cord ( Figure 2B) In keeping with the neuroinflammatory response that occurs in twitcher brain, we observed an increase in the levels of expression of the classical M1 macrophage polarization/microglia markers Fc receptor, IgG, low affinity III and II (Fcgr3/CD16 and Fcgr2/CD32) [29] with no changes in the expression of the alternative M2 macrophage polarization markers arginase-1 (Arg1) and mannose receptor C-type1 (Mrc1/CD206) [30]. A similar transcriptional profile was observed in the cerebellum ( Figure S1) and spinal cord ( Figure 2B) of Galc twi/twi mice that also showed an increase in the M2 polarization markers Arg1 and CD206. Notably, Ptx3 mRNA levels in the spinal cord of Galc twi/twi mice were significantly higher than those measured in the cerebrum and cerebellum counterparts at all time points investigated.
Ptx3 upregulation in the CNS of twitcher mice resulted in high levels of PTX3 protein in the cerebrum, cerebellum, and spinal cord of P35 twitcher mice when compared to control littermates as assessed by immunohistochemistry and Western blot analysis ( Figure 3A-C).
In addition, the levels of PTX3 in the blood of P35 twitcher mice were significantly increased when compared to Galc wt animals ( Figure 3D).
of Galc twi/twi mice that also showed an increase in the M2 polarization markers Arg1 and CD206. Notably, Ptx3 mRNA levels in the spinal cord of Galc twi/twi mice were significantly higher than those measured in the cerebrum and cerebellum counterparts at all time points investigated.
Ptx3 upregulation in the CNS of twitcher mice resulted in high levels of PTX3 protein in the cerebrum, cerebellum, and spinal cord of P35 twitcher mice when compared to control littermates as assessed by immunohistochemistry and Western blot analysis ( Figure  3A-C). In addition, the levels of PTX3 in the blood of P35 twitcher mice were significantly increased when compared to Galc wt animals ( Figure 3D). On this basis, immunohistochemical analysis was performed to identify PTX3-positive cells in the CNS of twitcher animals. As shown in Figure 4, double immunostains indicated that PTX3 immunoreactivity was prevalently detectable in IBA1-positive globoid cells in the white matter of the brain cortex and spinal cord of Galc twi/twi animals, no significant immunoreactivity being observed in control samples. In addition, PTX3 immunoreactivity was also present in GFAP-positive astrocytes of the gray matter of the CNS of On this basis, immunohistochemical analysis was performed to identify PTX3-positive cells in the CNS of twitcher animals. As shown in Figure 4, double immunostains indicated that PTX3 immunoreactivity was prevalently detectable in IBA1-positive globoid cells in the white matter of the brain cortex and spinal cord of Galc twi/twi animals, no significant immunoreactivity being observed in control samples. In addition, PTX3 immunoreactivity was also present in GFAP-positive astrocytes of the gray matter of the CNS of twitcher mice, but not of Galc wt animals that lacked signs of gliosis ( Figure 4). Similar results were obtained in twitcher cerebellum samples ( Figure S2).
At variance with the results obtained by the analysis of the twitcher CNS, no Ptx3 upregulation occur in the peripheral tissues of Galc twi/twi mice, including kidney, liver, and lungs, in which only limited psychosine accumulation and inflammatory responses occur because of GALC deficiency [23,25,[31][32][33] (Figure S3). Nevertheless, in keeping with the observation that the serum levels of PTX3 are increased after stroke [19,20] and in Parkinson's disease [21], the levels of PTX3 in the blood of P35 twitcher mice were significantly increased when compared to Galc wt animals ( Figure 3D). CNS of twitcher mice, but not of Galc wt animals that lacked signs of gliosis ( Figure 4). Similar results were obtained in twitcher cerebellum samples ( Figure S2). At variance with the results obtained by the analysis of the twitcher CNS, no Ptx3 upregulation occur in the peripheral tissues of Galc twi/twi mice, including kidney, liver, and lungs, in which only limited psychosine accumulation and inflammatory responses occur because of GALC deficiency [23,25,[31][32][33] (Figure S3). Nevertheless, in keeping with the observation that the serum levels of PTX3 are increased after stroke [19,20] and in Parkinson's disease [21], the levels of PTX3 in the blood of P35 twitcher mice were significantly increased when compared to Galc wt animals ( Figure 3D).

PTX3 Overexpression Reduces Clinical Symptoms and Spinal Cord Inflammation in Twitcher Mice
TgN(Tie2-hPTX3) mice (hPTX3 mice) ubiquitously express human PTX3 under the control of the Tie2 promoter, leading to a significant accumulation of the PTX3 protein in serum and tissues that occurs from birth throughout the whole life of the transgenic mice

PTX3 Overexpression Reduces Clinical Symptoms and Spinal Cord Inflammation in Twitcher Mice
TgN(Tie2-hPTX3) mice (hPTX3 mice) ubiquitously express human PTX3 under the control of the Tie2 promoter, leading to a significant accumulation of the PTX3 protein in serum and tissues that occurs from birth throughout the whole life of the transgenic mice [34]. The constitutive hPTX3 expression does not result in apparent defects in embryonic development; adult animals are normal and fertile, and no macroscopic or microscopic morphological abnormalities were observed in organs and tissues of hPTX3 mice [34].
On this basis, to assess a possible impact of the constitutive PTX3 overexpression in twitcher mice, syngeneic PTX3-overexpressing hPTX3 male mice were crossed with female Galc twi/+ animals to obtain hPTX3/Galc twi/+ breeders that were then crossed to generate hPTX3/Galc wt and hPTX3/Galc twi/twi animals. Next, hPTX3/Galc wt and hPTX3/Galc twi/twi mice were compared to the corresponding Galc wt and Galc twi/twi littermates. qPCR analysis using oligonucleotide primers designed to recognize simultaneously both human and murine PTX3 transcripts (Table S2) indicated that the levels of hPTX3 mRNA in the cerebrum of hPTX3/Galc wt mice were like the levels of the murine Ptx3 transcripts measured in Galc twi/twi mice at P35. In addition, hPTX3 overexpression together with the upregulation of its murine counterpart resulted in a cumulative increase in their mRNA levels in the brain of P35 hPTX3/Galc twi/twi mice when compared to Galc twi/twi littermates ( Figure S4A).
As shown in Figure S4C,D, constitutive hPTX3 overexpression did not affect the body weight gain of both hPTX3/Galc wt and hPTX3/Galc twi/twi mice when compared to the corresponding control Galc wt and Galc twi/twi animals, with Galc deficiency leading to a similar decrease in body weight starting from approximately day P20 and P28 both in the absence or in the presence of constitutive hPTX3 overexpression in male and female animals, respectively. Nevertheless, hPTX3 overexpression resulted in statistically significant, albeit limited 4-day increases in the overall survival of twitcher animals, extending their life span from 42 to 46 days (p < 0.001) ( Figure 5A), thus suggesting that PTX3 may exert an impact on the disease course. Indeed, when compared to Galc twi/twi animals, hPTX3/Galc twi/twi mice showed a reduced frequency and severity of twitching, which appeared around day P22 in both groups of animals ( Figure 5B,C). In addition, hPTX3/Galc twi/twi mice displayed a less severe atypical tail suspension reflex (hind limbs clenching) ( Figure 5D). No animal death and appearance of clinical symptoms were instead observed in hPTX3/Galc wt and Galc wt mice during the same investigation period.
Galc twi/+ animals to obtain hPTX3/Galc twi/+ breeders that were then crossed to generate hPTX3/Galc wt and hPTX3/Galc twi/twi animals. Next, hPTX3/Galc wt and hPTX3/Galc twi/twi mice were compared to the corresponding Galc wt and Galc twi/twi littermates. qPCR analysis using oligonucleotide primers designed to recognize simultaneously both human and murine PTX3 transcripts (Table S2) indicated that the levels of hPTX3 mRNA in the cerebrum of hPTX3/Galc wt mice were like the levels of the murine Ptx3 transcripts measured in Galc twi/twi mice at P35. In addition, hPTX3 overexpression together with the upregulation of its murine counterpart resulted in a cumulative increase in their mRNA levels in the brain of P35 hPTX3/Galc twi/twi mice when compared to Galc twi/twi littermates ( Figure S4A).
As shown in Figure S4C,D, constitutive hPTX3 overexpression did not affect the body weight gain of both hPTX3/Galc wt and hPTX3/Galc twi/twi mice when compared to the corresponding control Galc wt and Galc twi/twi animals, with Galc deficiency leading to a similar decrease in body weight starting from approximately day P20 and P28 both in the absence or in the presence of constitutive hPTX3 overexpression in male and female animals, respectively. Nevertheless, hPTX3 overexpression resulted in statistically significant, albeit limited 4-day increases in the overall survival of twitcher animals, extending their life span from 42 to 46 days (p < 0.001) ( Figure 5A), thus suggesting that PTX3 may exert an impact on the disease course. Indeed, when compared to Galc twi/twi animals, hPTX3/Galc twi/twi mice showed a reduced frequency and severity of twitching, which appeared around day P22 in both groups of animals ( Figure 5B,C). In addition, hPTX3/Galc twi/twi mice displayed a less severe atypical tail suspension reflex (hind limbs clenching) ( Figure 5D). No animal death and appearance of clinical symptoms were instead observed in hPTX3/Galc wt and Galc wt mice during the same investigation period.  and Galc twi/twi mice (n = 11). No neurological signs were observed before day P20 in both groups of animals. No animal death and clinical symptoms were observed in hPTX3/Galc wt and Galc wt mice throughout the whole experimental period. * p < 0.05; # p < 0.01; § p < 0.001, Student's t-test.
On this basis, we assessed the levels of expression of the microglial marker Iba1, of the astrocytic marker of gliosis Gfap, and of the myelin basic protein-encoding gene mbp in the cerebrum, cerebellum, and spinal cord of the different experimental groups at P35. As shown in Figure 6A, GALC deficiency caused a similar upregulation of Iba1 and Gfap genes and downregulation of mbp in the CNS of hPTX3/Galc twi/twi and Galc twi/twi mice at this late stage of the disease when compared to Galc wt and hPTX3/Galc wt animals. Accordingly, immunohistochemical analysis demonstrated that the IBA1-positive microglia infiltrate, GFAP-positive gliosis, and myelin degradation of MBP-positive fibers were present to a similar extent in the CNS of P35 hPTX3/Galc twi/twi and Galc twi/twi mice ( Figure 6B).  In keeping with the lack of a significant impact of constitutive hPTX3 overexpression on the histopathological features of the CNS of GALC-deficient animals, the levels of the neurotoxic GALC substrate psychosine were increased to a similar extent in the CNS of Galc twi/twi and hPTX3/Galc twi/twi animals ( Table 2). Finally, qPCR analysis of cerebrum, cerebellum, and spinal cord of P35 animals was performed to assess the impact of constitutive hPTX3 overexpression on the inflammatory response of the CNS in GALC-deficient mice. As shown in Figure 7, the expression of the inflammatory mediators Tnfα, Il1α, and Cxcl1 and of the M1 polarization macrophage/microglia markers CD16/CD32, albeit still higher than that observed in control wild type animals, was significantly reduced in the spinal cord of hPTX3/Galc twi/twi mice when compared to the Galc twi/twi counterpart, whereas Arg1 and CD206 expression remained unchanged.
In keeping with the lack of a significant impact of constitutive hPTX3 overexpre on the histopathological features of the CNS of GALC-deficient animals, the levels o neurotoxic GALC substrate psychosine were increased to a similar extent in the CN Galc twi/twi and hPTX3/Galc twi/twi animals ( Table 2). Finally, qPCR analysis of cerebrum, cerebellum, and spinal cord of P35 animals performed to assess the impact of constitutive hPTX3 overexpression on the inflamm response of the CNS in GALC-deficient mice. As shown in Figure 7, the expression o inflammatory mediators Tnfα, Il1α, and Cxcl1 and of the M1 polarization macrophag croglia markers CD16/CD32, albeit still higher than that observed in control wild typ imals, was significantly reduced in the spinal cord of hPTX3/Galc twi/twi mice when pared to the Galc twi/twi counterpart, whereas Arg1 and CD206 expression remained changed. Figure 7. qPCR analysis of the spinal cord of hPTX3/Galc wt and hPTX3/Galc twi/twi mice. Steady mRNA levels of the indicated genes were evaluated by qPCR in the spinal cord of hPTX3/Galc hPTX3/Galc twi/twi mice and compared to those measured in the corresponding control Galc w Galc twi/twi mice harvested at P35. Data were normalized to Gapdh expression and are the mean ± of 7-10 animals per group. n.s., not significant; * p < 0.1; ** p < 0.05; *** p < 0.01, **** p < 0.0 Galc twi/twi versus hPTX3/Galc twi/twi mice; §, p < 0.001 for Galc wt versus Galc twi/twi mice or for hPTX3/ versus hPTX3/Galc twi/twi animals. One-way ANOVA with post hoc comparisons with adjustme multiple comparisons (Sidak).
At variance, no differences in gene expression were detected in the cerebrum cerebellum of the two groups of animals, the only exceptions being represented by a s increase in the expression of TNFα in the cerebrum of hPTX3/Galc twi/twi mice when pared to Galc twi/twi animals and a decreased cerebellar expression of the M2 polariz markers Arg1 and CD206 that occurred also in hPTX3/Galc wt mice, possibly due t PTX3 upregulation present in both groups ( Figure S5). In keeping with these observat the levels of transcription factor Cebpd were significantly reduced in the spinal co Figure 7. qPCR analysis of the spinal cord of hPTX3/Galc wt and hPTX3/Galc twi/twi mice. Steady state mRNA levels of the indicated genes were evaluated by qPCR in the spinal cord of hPTX3/Galc wt and hPTX3/Galc twi/twi mice and compared to those measured in the corresponding control Galc wt and Galc twi/twi mice harvested at P35. Data were normalized to Gapdh expression and are the mean ± SEM of 7-10 animals per group. n.s., not significant; * p < 0.1; ** p < 0.05; *** p < 0.01, **** p < 0.001 for Galc twi/twi versus hPTX3/Galc twi/twi mice; §, p < 0.001 for Galc wt versus Galc twi/twi mice or for hPTX3/Galc wt versus hPTX3/Galc twi/twi animals. One-way ANOVA with post hoc comparisons with adjustment for multiple comparisons (Sidak).
At variance, no differences in gene expression were detected in the cerebrum and cerebellum of the two groups of animals, the only exceptions being represented by a slight increase in the expression of TNFα in the cerebrum of hPTX3/Galc twi/twi mice when compared to Galc twi/twi animals and a decreased cerebellar expression of the M2 polarization markers Arg1 and CD206 that occurred also in hPTX3/Galc wt mice, possibly due to the PTX3 upregulation present in both groups ( Figure S5). In keeping with these observations, the levels of transcription factor Cebpd were significantly reduced in the spinal cord of hPTX3/Galc twi/twi mice when compared to Galc twi/twi animals (Figure 7), no differences being instead observed between the two groups in the cerebrum and cerebellum ( Figure S5). Accordingly, the levels of murine Ptx3 transcript were reduced only in the spinal cord of hPTX3/Galc twi/twi animals ( Figure 7). Thus, at variance with what was observed in the cerebrum (see above), the cumulative levels of the human plus murine PTX3 transcripts were reduced in the spinal cord of hPTX3/Galc twi/twi animals relative to the levels of murine Ptx3 mRNA in Galc twi/twi mice (see Figure S4B).

Discussion
In humans, genetic deficiency of the sphingolipid-metabolizing enzyme GALC leads to Krabbe disease, a neuroinflammatory degenerative disorder. Here, we demonstrated that the soluble pattern recognition receptor PTX3 was expressed by monocyte-derived cells in brain specimens from Krabbe patients, its immunoreactivity being related to the extent of the pathology and gliosis. In keeping with these observations, twitcher mice, an authentic model of Krabbe disease [22,23], were characterized by the progressive upregulation of PTX3 along the CNS caudal-rostral axis, with no changes in the peripheral organs. The upregulation of the Ptx3 transcript levels in cerebrum, cerebellum, and spinal cord of twitcher mice resulted in an increase in PTX3 protein levels in the affected tissues and plasma.
Our data extend these observations to Krabbe patients and demonstrate for the first time that Ptx3 upregulation occurs in the CNS of twitcher mice in parallel with an increased expression of neuroinflammation-related genes, including various cytokines/chemokines, the marker of gliosis Gfap, the microglial marker Iba1, and the M1 macrophage polarization/microglia markers CD16/CD32. These changes were detectable at postnatal day P15, before the onset of evident neurologic signs, and were increased at P24, when pathologic alterations were occasionally noticeable, and at P35, when clear neurologic defects occurred, including tremor and hind limb paralysis [23,25,26]. These findings indicate that the inflammatory environment progressively established after birth in the CNS of GALCdeficient mice drives the expression of Ptx3, as confirmed by the observed upregulation of Cebpd, a transcription factor known to mediate PTX3 expression in astrocytes during neuroinflammation [27].
Previous findings had shown that PTX3 can be expressed by neurons, astrocytes, and/or microglia following cytokine stimulation or under neuroinflammatory conditions, depending on the specific disorder and acute versus chronic phase of the disease (see [15,35,36] and references therein). Here, immunohistochemical analysis demonstrated that autopsy brain specimens from Krabbe infants were characterized by an intense gliosis and PTX3 immunoreactivity that was detectable in macrophages and globoid cells, a hallmark of Krabbe disease [9]. These observations were confirmed by the analysis of Galc twi/twi mice, in which PTX3 immunoreactivity was prevalently detectable in IBA1-positive globoid cells of the white matter of the brain cortex, cerebellum, and spinal cord of these animals, as well as in GFAP-positive astrocytes of the gray matter.
PTX3 is an important mediator of innate immune responses, produced locally at the site of inflammation [37]. In CNS, PTX3 modulates the activity of microglia by inhibiting phagocytosis of apoptotic cells and favoring the uptake of pathogens [24]. PTX3 binding helps the rescue of neurons from phagocytic clearance by macrophages [27] and protects them from ischemic damage [19,38], trauma [39], and in Parkinson's disease [40]. Accordingly, neurogenesis and angiogenesis were inhibited after cerebral ischemia in Ptx3 null mice, and neuronal damage was increased after limbic seizure when Ptx3 deficient animals were compared to controls [17,18]. Together, these data point to a protective role of PTX3 during neuroinflammation.
Our data demonstrate that the constitutive overexpression of hPTX3 that occurs from birth throughout the whole life of hPTX3/Galc twi/twi mice attenuates the severity of clinical signs, such as twitching and hind limbs clenching, and causes a significant, albeit limited, increase in their life span. However, when assessed at the late P35 stage of the disease, hPTX3 overexpression did not exert any significant effect on the CNS of GALC-deficient mice in terms of IBA1-positive microglia infiltrate, GFAP-positive gliosis, and myelin degradation of MBP-positive fibers. In addition, a similar accumulation of the neurotoxic GALC substrate psychosine was observed in the cerebrum and spinal cord of Galc twi/twi and hPTX3/Galc twi/twi mice at this time point. These data appear to be in keeping with previous observations indicating that PTX3 is unable to exert a significant impact on the histopathological damage that occurs in mice after experimental neurotrauma or to affect the course of experimental autoimmune encephalomyelitis [35,36]. Nevertheless, we found that hPTX3 overexpression reduced the upregulation of proinflammatory genes in the spinal cord of P35 hPTX3/Galc twi/twi mice when compared to Galc twi/twi animals, including endogenous Ptx3 and its transcription factor Cebpd, with no effect on the neuroinflammatory response observed in the cerebrum and cerebellum of these animals. Relevant to this point, neurohistopathological and neurochemical alterations caused by GALC deficiency in the CNS of twitcher mice arise in a temporal and region-dependent fashion, the first alterations being observed in the spinal cord to progress along the caudal-rostral axis in the cerebellum and cerebrum (see [41] and references therein). In addition, spinal cord alterations were already observed in 18-21-week-old GLD fetuses [42,43]. Thus, the spinal cord represents the earliest tissue affected by GALC deficiency in murine and human CNS and appears to be more prone to a protective effect exerted by PTX3 on the neuroinflammatory response that occurs in twitcher mice.
Different hypotheses can be raised to attempt to explain why the spinal cord is more responsive to the partial protecting role exerted by genetic, constitutive hPTX3 upregulation in GALC-deficient mice. The ubiquitous production of PTX3 in hPTX3 animals is under the control of the Tie2 promoter [34] that drives the vascular expression of the transgene throughout embryogenesis and adulthood [44]. Thus, the early and long-lasting PTX3 overexpression that occurs in hPTX3/Galc twi/twi animals may prevent, at least in part, the first lesions that arise in the spinal cord of GALC-deficient mice, being instead less effective against the later CNS lesions (further experiments performed on hPTX3/Galc twi/twi mice at earlier stages of the disease will be required to elucidate this point). In addition, the metabolic alterations that occur in parallel with the progression of gliosis, neurodegeneration, microglial activation, and apoptosis along the rostral-caudal axis in a regional and age-dependent fashion [41] may exert a different impact on the protective activity of PTX3 in hPTX3/Galc twi/twi mice. A further hypothesis can be based on the fact that PTX3 may also exert detrimental effects following CNS damage (see [15] and references therein), indicating that the activity of PTX3 in neuroinflammation may represent the result of a fine tuning in different areas of the CNS between pro-and anti-neurodegenerative mechanisms of action that remain poorly defined. Finally, other members of the long pentraxin family, represented by the neuronal pentraxins NP1, NP2 and NPR, may exert an impact on neurodegeneration and interfere with PTX3 activity [35]. Experiments performed in a Ptx3 null background would be useful to assess these hypotheses. However, due to the female infertility of Ptx3 null mice [45] and the early lethality of Galc twi/twi animals, the generation of double Ptx3 −/− /Galc twi/twi mice would require a breeding program with a very large number of animals, incompatible with the rules of the local and national ethical committees.
The neurotoxic GALC substrate psychosine has been used as a biomarker to identify patients with infantile GLD, and possibly late onset GLD patients, and to monitor the timing and efficacy of hematopoietic stem cell transplantation in these patients [46]. Our data demonstrate that serum levels of PTX3 increase in twitcher mice. Further studies will be required to assess whether longitudinal blood measurement of PTX3 levels may represent a novel biomarker more sensitive than psychosine to predict onset after newborn screening and to follow disease progression in transplanted and/or late-onset GLD patients.
In conclusion, our results provide the first evidence that PTX3 is produced in the CNS of Krabbe patients and GALC-deficient twitcher mice and may exert a partial protective role by reducing the neuroinflammatory response that occurs in the spinal cord of Galc twi/twi animals, with possible implications for GLD management and therapy.

Histopathology of Human GLD Biopsies
Autopsy brain specimens of GLD patients (Table S1)  Formalin-fixed, paraffin-embedded tissue sections were submitted to H&E and single or double immunohistochemical staining. Briefly, sections were de-waxed, rehydrated, and endogenous peroxidase activity blocked with 0.3% H 2 O 2 in methanol for 20 min. Antigen retrieval was performed using a microwave oven or thermostatic bath in 1.0 mM EDTA buffer (pH 8.0). Sections were then washed in TBS (pH 7.4) and incubated for 1 h with the specific primary antibody diluted in TBS 1% bovine serum albumin. Signal was revealed using the DAKO Envision+System-HRP Labelled Polymer Anti-Mouse or Anti-Rabbit (Dako Cytomation, Santa Clara, CA, USA), followed by Diaminobenzydine (DAB) as chromogen and hematoxylin as counterstain.
For double immunostains, after completing the first immune reaction, the second primary antibody was applied and labelled using MACH 4 TM Universal AP Polymer Kit (Biocare Medical, Pacheco, CA, USA); chromogen reaction was developed with Ferangi Blue TM Chromogen System (Biocare Medical), and nuclei were faintly counterstained with Methyl Green. Images were then acquired with an Olympus DP70 camera mounted on an Olympus Bx60 microscope using AnalySIS imaging software (Soft Imaging System GmbH, Münster, Germany). The following primary antibodies were used: rabbit anti-PTX3 polyclonal antibody (1:500, kindly provided by B. Bottazzi Humanitas Clinical Institute, Rozzano, Italy), monoclonal mouse anti human CD68, clone PG-M1 (1:200, Dako Cytomation), and monoclonal mouse anti human GFAP, clone 6F2 (1:100, Dako Cytomation).

Animals
Breeder twitcher heterozygous mice (C57BL/6J, Galc twi/+ ; Jackson Laboratories, Bar Harbor, ME, USA) were maintained under standard housing conditions. Experiments were performed according to the Italian laws (D.L. 116/92 and following additions) that enforce the EU 86/109 Directive and were approved by the local animal ethics committee (OPBA, Università degli Studi di Brescia, Italy). Twitcher mutation was determined by polymerase chain reaction (PCR) on DNA extracted from clipped tails [47]. In all of the experiments, littermate wild type (Galc wt ) and homozygous (Galc twi/twi ) animals were used. To generate PTX3-overexpressing twitcher mice, syngeneic TgN(Tie2-hPTX3) male mice (hPTX3 mice) that ubiquitously express human PTX3 under the control of the Tie2 promoter [34] were bred with female Galc twi/+ mice to obtain hPTX3/Galc twi/+ animals. Next, hPTX3/Galc wt and hPTX3/Galc twi/twi mice were generated by crossing hPTX3/Galc twi/+ breeder mice that were genotyped for Galc status by PCR and for hPTX3 overexpression by RT-PCR. For the survival studies, the end point was established for each mouse by the animal house veterinarian unaware of the animal genotype, according to the rules of the local ethics committee.

Quantitative RT-PCR Analysis
Cerebrum, cerebellum, and spinal cord specimens were analyzed for gene expression by quantitative RT-PCR (qPCR) at P15, P24 and/or P35, and data were normalized for Gapdh expression [48]. For this purpose, total RNA was extracted from frozen samples using TRIzol Reagent according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA), and contaminating DNA was digested using DNAse (Promega, Madison, WI, USA). Total RNA (2 µg) was retrotranscribed with MMLV reverse transcriptase (Invitrogen) using random hexaprimers in a final 20 µL volume. Quantitative PCR was performed with a ViiA TM 7 Real-Time PCR Detection System (Applied Biosystems, Waltham, MA, USA) using an iQTM SYBR Green Supermix (Biorad, Hercules, CA, USA) according to the manufacturer's instructions. Ptx3 expression levels were analyzed by qPCR also in kidneys, liver, and lungs from Galc wt and Galc twi/twi mice at P35. In each experiment, an arbitrary value equal to 1.0 was assigned to the levels of expression of the gene(s) measured in one sample that was used as reference. The specific primers are shown in Table S2.
For double immunostains, incubation with the anti-PTX3 polyclonal antibody was followed by 1 h incubation with biotin anti-rabbit antibody and by 1 h incubation with streptavidin Alexa Fluor 594. Sections were then incubated for 2 h with anti-GFAP or anti-Iba1 monoclonal antibody, followed by 1 h incubation with Alexa Fluor 488 anti-mouse antibody. All tissue sections were incubated for 30 min with DAPI for nuclear staining and mounted in Dako fluorescent mounting medium.
Images were taken with an Axiovert 200 M microscope equipped with ApoTome optical sectioning device (Carl Zeiss, Oberkochen, Germany) using the same settings for comparison of all samples.

Assessment of Animal Clinical Features
Body weight, life span, twitching and hind limbs clenching were monitored daily from day P8 until mice reached a moribund condition. Each mouse was observed for at least 1 min by two trained observers. The extent of frequency and severity of twitching was scored as: 1, fine; 2, mild; 3, mild-moderate; 4, moderate; 5, severe. Hind limbs clenching frequency was scored: 1, rare; 2, mild; 3, intermittent; 4, moderate; 5, severe [49].

Psychosine Extraction and MS Analysis
Psychosine extraction was performed as previously described [50]. Briefly, 20 mg of lyophilized brain or spinal cord were extracted with 15 mL of acetone using a Polytron homogenizer to selectively extract psychosine. The extract was filtered through a Büchner funnel and evaporated to dryness. The dried residue was dissolved in 2.0 mL chloroform/methanol (2/1 v/v), and the psychosine-enriched fraction was evaporated to dryness, dissolved in methanol, and further analyzed by mass spectrometry (MS).
All MS analyses were performed at the Unitech OMICs platform (University of Milano, Italy) using an ExionLC™ AD system connected to TripleTOF™ 6600 System equipped with Turbo V™ Ion Source with ESI Probe (SCIEX, Framingham, MA, USA). Samples were separated on a Kinetex ® EVO C18 100 (Length) × 2.1 mm (ID) × 1.7 µm (particle size). The temperature was set at 40 • C. The analytes were eluted with the following gradient: from 60% buffer A (0.01% formic acid in water) to 99% buffer B (0.01% formic acid in methanol) in 7.5 min. Constant flow rate: 450 µL/min. Total run: 12 min. MS spectra were collected, in positive polarity, in full-mass scan from 250 to 800 Da (100 ms accumulation time) and in IDA ® mode (information-dependent acquisition) from 100 to 800 Da (40 ms accumulation time, top 18 spectra per cycle 0.87 s). Nitrogen was used as a nebulizing gas (GS1, 55 psi), turbo spray gas (GS2, 65 psi), and curtain gas (CUR, 35 psi). Spray voltage was fixed at 5.0 kV, de-clustering potential (DP) was 50 eV, the collision energy was 30 eV with a collision energy spread (CES) of 15 eV, and source temperature was 300 • C.

Statistical Analysis
Comparisons among multiple groups were performed using a one-way ANOVA model, followed by post hoc comparisons with adjustment for multiple comparisons (Sidak procedure). Comparison between two groups was performed using the Student's unpaired t-test.