This study was the first to describe the IHC findings of BBB in CNS tissue specimens of NPCM and HIV/NPCM participants. This study observed a marked decrease in CLDN-5 expression in the capillaries of the HIV/NPCM coinfected tissues (acute/subacute clinical form of PCM) compared with that in the HIV or NPCM monoinfected (chronic clinical form of PCM) tissues. This is in accordance with the stated hypothesis that there would be an additive effect on BBB disruption with HIV/NPCM co-infection compared with monoinfection. However, additional studies with more CNS specimens are warranted to prove this hypothesis.
A higher expression of CLDN-1 in the NPCM and NPCM/HIV coinfected tissues, but not in the HIV monoinfected tissues, was observed compared with that in the normal control tissues. This is in accordance with the observation that CLDN-1 is rarely expressed at the normal BBB, although it is expressed in pathological conditions. The CLDN-1 expression, post stroke, results in limited brain endothelial barrier recovery, partially owing to restricted CLDN-5 expression . The increased expression of CLDN-1 in the CNS vasculature of the NPCM/HIV tissues described in the current study is probably compensatory to the decrease of CLDN-5 in the same patient. Some studies question if CLDN-5 deficiency is compensated with other TJ proteins. For others, an interplay between TJ proteins is considered a pre-requisite for the regulation of barrier permeability and for compensatory mechanisms, but little about this is known . At the brain endothelium, CLDN-5 is the major claudin, but there are conditions (e.g., vasculo genesis, tumors, and traumatic brain injury) in which other claudins such as CLDN-1 are present and might participate in paracellular space occlusion [42, 43]. It was found that there was significant upregulation of CLDN-1 mRNA protein and a nonspecific CLDN in blood vessels and downregulation in CLDN-5 expression .
Redistribution of transmembrane junction proteins, particularly CLDN-5 and occludin as well as Ve-cadherin, is closely associated with loosening of adhesion at the TJ complex, which ultimately leads to paracellular gap formation .
According to the present study, in the normal control brain tissue specimens, the expression of CLDN-1 was lower than that of CLDN-3 or -5. This is in accordance with the hypothesis that CLDN-1 is not essential for BBB TJ complex function under physiological conditions and its induction is rather associated with an altered TJ function. The cause and effects of CLDN-1 expression are still unknown. Some studies suggested an equal and beneficial role in BBB TJ complex organization, helping to seal the paracellular route, or via astrocytic expression, helping to form secondary astrocyte/glial barriers to protect the brain from an endothelial breakdown [42, 44]. In this study, CLDN-1 expression was highly expressed in NPCM monoinfected tissues, which indicates a chronic form of PCM in the spinal cord. This is explained by the fact that the chronic form of PCM presented with a longer clinical duration than the acute/subacute form in HIV/NPCM coinfected cases. The appearance of CLDN-1, chronically in stroke, raises the possibility that it may support and facilitate BBB recovery, reestablishing integrity, as indicated in the study on the beneficial effects of overexpressing CLDN-1 in experimental autoimmune encephalomyelitis .
Disruption of TJ is commonly observed in HIV-1-infected patients . The HIV-1 and its proteins may act alone or in cooperation with host cytokines and chemokines to affect the integrity of the BBB .
Tat can affect protein expression and distribution of TJ; such changes may lead to disturbances in BBB integrity and allow viral flow to the brain . In the present study, the expression of CLDN-1 was lower in HIV monoinfected tissues than in NPCM/HIV coinfected tissues. It was described that the Tat protein caused a reduction of CLDN-1 immunoreactivity and loss of CLDN-5 expression, resulting in marked cellular infiltration . Changes in CLDN-5 immunoreactivity were associated with increased cells in the brain tissue, suggesting that monocytes may migrate to the brain through tight juncture disruption caused by CLDN loss . Tat also disrupts the integrity of BBB by diminishing OCLN production and cleaving OCLN through matrix MMP-9 [16, 48]. HIV-1 Tat also affects AJ proteins; it suppresses β-catenin activity in astrocytes by inhibiting microRNAs .It was reported that gp120 induced disruption of the BBB via brain microvessels injury, MMP activation, and degradation of the vascular basement membrane and vascular TJ . Gp120 proteins caused disruption and downregulation of the TJ proteins ZO-1, ZO-2, and OCLN in these cells. Other junctional proteins such as CLDN-1 and CLDN-5 were unaffected by gp120 treatment. These data demonstrate that the gp120 proteins alter both the functional and molecular properties of the BBB, which could increase trafficking of HIV, infected cells, and toxic humoral factors into the CNS and contribute to the pathogenesis of HIV-associated dementia (HAD) . Moreover, gp-120 proteins increased the expression of MMP-2 and − 9, and subsequently that of CLDN- 5 and laminin, which are key targets of these MMPs . A role for gp120 in BBB disruption is also supported by in vitro evidence that gp120 added to cultured human brain endothelial cells enhances monocyte migration, increases permeability, decreases TEER, and disrupts expression of TJ proteins ZO-1, ZO-2, and OCLN (but not CLDNs or actin) [31, 52].The VPR is a 96-amino acid 14-kDa protein. It has a regulatory function of viral transcription, plays an important role in regulating nuclear import of the HIV-1 pre-integration complex, and is required for viral replication in non-dividing cells such as macrophages. High levels of VPR in the CSF are found in patients with HAD. It induces apoptosis in neurons via caspase activation . Other HIV proteins, such as nef and regulator of expression of virion (rev) proteins, also play a role in the rupture of BBB, inducing neurotoxicity . A segment in Nef protein contains identical amino acids at key positions and structurally mimics the β-catenin binding sites on endogenous β-catenin ligands. The interaction between Nef and β-catenin was confirmed with in vitro studies .
The main limitations of this study were its descriptive design and the few numbers of brain tissue specimens studied. The scarce number of CNS specimens from autopsy or biopsy could explain this paucity of information. In the present study, it was possible to localize only two FFPE CNS tissue specimens out of 10 identified with NPCM, in the Anatomic Pathology Service of a University Hospital, localized in an endemic area, in a period of 64 years . To date, all five reported cases of NPCM in HIV-positive patients are from Brazil [3, 57–60]. This study was performed with FFPE; for which, the brain tissue was stored in 30% formaldehyde during the 30 days before the brain macroscopic study. This hindered antigen recuperation, which takes 60 minutes, resulting in the development of a strong background on IHC. Besides this, there were marked differences between the NPCM monoinfected and NPCM/HIV coinfected tissue specimens. In NPCM, the clinical form was chronic and the second was acute/subacute, which implies differences in disease duration. The NPCM monoinfected tissue specimen was from a previously immunocompetent patient, whereas that of NPCM/HIV was from a severely immunosuppressed patient, with multiple CNS opportunistic infections, such as toxoplasmosis, histoplasmosis and candidiasis . The NPCM/HIV specimen was that of an autopsy case, and the NPCM monoinfected specimen was obtained following spinal cord biopsy (alive). These differences can cause an impressive impact on the BBB. NPCM/HIV, besides the NPCM, presented at autopsy, multiple opportunistic infections. These differences could be responsible for the differences in the level of the CLDN studied.
This study has several strengths as well. TJ and AJ proteins were studied for the first time in NPCM, contributing to the pathophysiology of CNS PCM infection and HIV co-infection. In conclusion, there was a marked decrease in CLDN-5 expression in the NPCM/HIV coinfected tissue specimen, with compensatory increase of CLDN-1, probably different from NPCM or HIV monoinfected cases.
The mechanism by which Paracoccidioides sp. invades the CNS is unknown. Probably Paracoccidioides sp. invades the CNS using any of the two mechanisms (Fig. 4): (a) by using macrophages (Trojan horse), as Paracoccidioides sp. was identified inside macrophages  or (b) by using the paracellular pathway, mainly owing to the alteration of CLDN expression, which could be enhanced by HIV infection. However, additional studies are necessary to better explore this subject.