Comprehensive bioinformation analysis of homeodomain-leucine zipper gene family and expression pattern of HD-Zip I under abiotic stress in Salix suchowensis.

BACKGROUND: Homeodomain-leucine zipper (HD-Zip) transcription factors are plant-specific and play important roles in plant defense against environmental stresses. Identification and functional studies have been carried out in model plants such as rice, Arabidopsis thaliana, and poplar, but comprehensive analysis on the HD-Zip family of Salix suchowensis have not been reported. RESULTS: A total of 55 HD-Zip genes were identified in the willow genome, unevenly distributed on 18 chromosomes except for chromosome 19. And segmental duplication events containing SsHD-Zip were detected on all chromosomes except chromosomes 13 and 19. The SsHD-Zip were classified into 4 subfamilies subfamilies (I-IV) according to the evolutionary analysis, and members of each subfamily shared similar domain structure and gene structure. The combination of GO annotation and promoter analysis showed that SsHD-Zip genes responded to multiple abiotic stresses. Furthermore, the results of qPCR analysis showed that the SsHD-Zip I gene exhibited different degrees of expression under salt stress, PEG treatment and heat treatment. Moreover, there was a synergistic effect between SsHD-Zip I genes under stress conditions based on coregulatory networks analysis. CONCLUSIONS: In this study, HD-Zip transcription factors were systematically identified and analyzed at the whole genome level. These results preliminarily clarified the structural characteristics and related functions of willow HD-Zip family members, and it was found that SsHox34, SsHox36 and SsHox51 genes were significantly involved in the response to various stresses. Together, these findings laid the foundation for further research on the resistance functions of willow HD-Zip genes.

HD-Zip IV transcription factors: Drivers of epidermal cell fate integrate metabolic signals.

The leaf epidermis comprises the outermost layer of cells that protect plants against environmental stresses such as drought, ultraviolet radiation, and pathogen attack. Research over the past decades highlights the role of class IV homeodomain leucine-zipper (HD-Zip IV) transcription factors (TFs) in driving differentiation of various epidermal cell types, such as trichomes, guard cells, and pavement cells. Evolutionary origins of this family in the charophycean green algae and HD-Zip-specific gene expression in the maternal genome provide clues to unlocking their secrets which include ties to cell cycle regulation. A distinguishing feature of these TFs is the presence of a lipid binding pocket that integrates metabolic information with gene expression. Identities of metabolic partners are beginning to emerge, uncovering feedback loops to maintain epidermal cell specification. Discoveries of associated molecular mechanisms are revealing fascinating links to phospholipid and sphingolipid metabolism and mechanical signaling.

Rapid Evolution of the Embryonically Expressed Homeobox Gene LEUTX within Primates.

LEUTX is a homeodomain transcription factor expressed in the very early embryo with a function around embryonic genome activation. The LEUTX gene is found only in eutherian mammals including humans but, unlike the majority of homeobox genes, the encoded amino acid sequence is very different between divergent mammalian species. However, whether dynamic evolution has also occurred between closely related mammalian species remains unclear. In this work, we perform a comparative genomics study of LEUTX within the primates, revealing dramatic evolutionary sequence change between closely related species. Positive selection has acted on sites in the LEUTX protein, including six sites within the homeodomain; this suggests that selection has driven changes in the set of downstream targets. Transfection into cell culture followed by transcriptomic analysis reveals small functional differences between human and marmoset LEUTX, suggesting rapid sequence evolution has fine-tuned the role of this homeodomain protein within the primates.

Negatively charged, intrinsically disordered regions can accelerate target search by DNA-binding proteins.

In eukaryotes, many DNA/RNA-binding proteins possess intrinsically disordered regions (IDRs) with large negative charge, some of which involve a consecutive sequence of aspartate (D) or glutamate (E) residues. We refer to them as D/E repeats. The functional role of D/E repeats is not well understood, though some of them are known to cause autoinhibition through intramolecular electrostatic interaction with functional domains. In this work, we investigated the impacts of D/E repeats on the target DNA search kinetics for the high-mobility group box 1 (HMGB1) protein and the artificial protein constructs of the Antp homeodomain fused with D/E repeats of varied lengths. Our experimental data showed that D/E repeats of particular lengths can accelerate the target association in the overwhelming presence of non-functional high-affinity ligands ('decoys'). Our coarse-grained molecular dynamics (CGMD) simulations showed that the autoinhibited proteins can bind to DNA and transition into the uninhibited complex with DNA through an electrostatically driven induced-fit process. In conjunction with the CGMD simulations, our kinetic model can explain how D/E repeats can accelerate the target association process in the presence of decoys. This study illuminates an unprecedented role of the negatively charged IDRs in the target search process.

Many DNA/RNA-binding proteins possess intrinsically disordered regions (IDRs) with large negative charge, some of which involve a consecutive sequence of aspartate (D) or glutamate (E) residues. We refer to them as D/E repeats. The functional role of D/E repeats is not well understood, though some of them are known to cause autoinhibition. Here, using the HMGB1 protein and the artificial protein constructs of the Antp homeodomain fused with D/E repeats, we demonstrate that D/E repeats can accelerate the target search process in the presence of non-functional high-affinity ligands ('decoys'). Our coarse-grained molecular dynamics (CGMD) simulations and kinetic model provide mechanistic insight into this acceleration. Our current study illuminates an unprecedented role of the negatively charged IDRs.

A complete methyl-lysine binding aromatic cage constructed by two domains of PHF2.

The N-terminal half of PHF2 harbors both a plant homeodomain (PHD) and a Jumonji domain. The PHD recognizes both histone H3 trimethylated at lysine 4 and methylated nonhistone proteins including vaccinia-related kinase 1 (VRK1). The Jumonji domain erases the repressive dimethylation mark from histone H3 lysine 9 (H3K9me2) at select promoters. The N-terminal amino acid sequences of H3 (AR2TK4) and VRK1 (PR2VK4) bear an arginine at position 2 and lysine at position 4. Here, we show that the PHF2 N-terminal half binds to H3 and VRK1 peptides containing K4me3, with dissociation constants (KD values) of 160 nM and 42 nM, respectively, which are 4 × and 21 × lower (and higher affinities) than for the isolated PHD domain of PHF2. X-ray crystallography revealed that the K4me3-containing peptide is positioned within the PHD and Jumonji interface, with the positively charged R2 residue engaging acidic residues of the PHD and Jumonji domains and with the K4me3 moiety encircled by aromatic residues from both domains. We suggest that the micromolar binding affinities commonly observed for isolated methyl-lysine reader domains could be improved via additional functional interactions within the same polypeptide or its binding partners.

Anionic lipids induce a fold-unfold transition in the membrane-translocating Engrailed homeodomain.

Homeoprotein transcription factors have the property of interacting with membranes through their DNA-binding homeodomain, which is involved in unconventional internalization and secretion. Both processes depend on membrane-translocating events but their detailed molecular mechanisms are still poorly understood. We have previously characterized the conformational properties of Engrailed 2 homeodomain (EnHD) in aqueous solution and in micelles as membrane-mimetic environments. In the present study, we used small isotropic lipid bicelles as a more relevant membrane-mimetic model to characterize the membrane-bound state of EnHD. We show that lipid bicelles, in contrast to micelles, adequately reproduce the requirement of anionic lipids in the membrane binding and conformational transition of EnHD. The fold-unfold transition of EnHD induced by anionic lipids was characterized by NMR using 1H, 13C, 15N chemical shifts, nuclear Overhauser effects, residual dipolar couplings, intramolecular and intermolecular paramagnetic relaxation enhancements induced by site-directed spin-label or paramagnetic lipid probe, respectively. A global unpacking of EnHD helices is observed leading to a loss of the native fold. However, near-native propensities of EnHD backbone conformation are maintained in membrane environment, including not only the three helices but also the turn connecting helices H2 and H3. NMR and coarse-grained molecular dynamics simulations reveal that the EnHD adopts a shallow insertion in the membrane, with the three helices oriented parallel to the membrane. EnHD explores extended conformations and closed U-shaped conformations, which are stabilized by anionic lipid recruitment.

Elucidating the importance and regulation of key enhancers for human MEIS1 expression.

Myeloid ecotropic virus insertion site 1 (MEIS1) is essential for normal hematopoiesis and is a critical factor in the pathogenesis of a large subset of acute myeloid leukemia (AML). Despite the clinical relevance of MEIS1, its regulation is largely unknown. To understand the transcriptional regulatory mechanisms contributing to human MEIS1 expression, we created a knock-in green florescent protein (GFP) reporter system at the endogenous MEIS1 locus in a human AML cell line. Using this model, we have delineated and dissected a critical enhancer region of the MEIS1 locus for transcription factor (TF) binding through in silico prediction in combination with oligo pull-down, mass-spectrometry and knockout analysis leading to the identification of FLI1, an E-twenty-six (ETS) transcription factor, as an important regulator of MEIS1 transcription. We further show direct binding of FLI1 to the MEIS1 locus in human AML cell lines as well as enrichment of histone acetylation in MEIS1-high healthy and leukemic cells. We also observe a positive correlation between high FLI1 transcript levels and worse overall survival in AML patients. Our study expands the role of ETS factors in AML and our model constitutes a feasible tool for a more detailed understanding of transcriptional regulatory elements and their interactome.

Proximity labeling identifies a repertoire of site-specific R-loop modulators.

R-loops are three-stranded nucleic acid structures that accumulate on chromatin in neurological diseases and cancers and contribute to genome instability. Using a proximity-dependent labeling system, we identified distinct classes of proteins that regulate R-loops in vivo through different mechanisms. We show that ATRX suppresses R-loops by interacting with RNAs and preventing R-loop formation. Our proteomics screen also discovered an unexpected enrichment for proteins containing zinc fingers and homeodomains. One of the most consistently enriched proteins was activity-dependent neuroprotective protein (ADNP), which is frequently mutated in ASD and causal in ADNP syndrome. We find that ADNP resolves R-loops in vitro and that it is necessary to suppress R-loops in vivo at its genomic targets. Furthermore, deletion of the ADNP homeodomain severely diminishes R-loop resolution activity in vitro, results in R-loop accumulation at ADNP targets, and compromises neuronal differentiation. Notably, patient-derived human induced pluripotent stem cells that contain an ADNP syndrome-causing mutation exhibit R-loop and CTCF accumulation at ADNP targets. Our findings point to a specific role for ADNP-mediated R-loop resolution in physiological and pathological neuronal function and, more broadly, to a role for zinc finger and homeodomain proteins in R-loop regulation, with important implications for developmental disorders and cancers.

Identification of a modular super-enhancer in murine retinal development.

Super-enhancers are expansive regions of genomic DNA comprised of multiple putative enhancers that contribute to the dynamic gene expression patterns during development. This is particularly important in neurogenesis because many essential transcription factors have complex developmental stage- and cell-type specific expression patterns across the central nervous system. In the developing retina, Vsx2 is expressed in retinal progenitor cells and is maintained in differentiated bipolar neurons and Müller glia. A single super-enhancer controls this complex and dynamic pattern of expression. Here we show that deletion of one region disrupts retinal progenitor cell proliferation but does not affect cell fate specification. The deletion of another region has no effect on retinal progenitor cell proliferation but instead leads to a complete loss of bipolar neurons. This prototypical super-enhancer may serve as a model for dissecting the complex gene expression patterns for neurogenic transcription factors during development. Moreover, it provides a unique opportunity to alter expression of individual transcription factors in particular cell types at specific stages of development. This provides a deeper understanding of function that cannot be achieved with traditional knockout mouse approaches.

Secondary structures, dynamics, and DNA binding of the homeodomain of human SIX1.

Human sine oculis homeobox homolog (SIX) 1 contains a homeodomain (HD), which is important for binding to DNA. In this study, we carried out structural studies on the HD of human SIX1 using nuclear magnetic resonance (NMR) spectroscopy. Its secondary structures and dynamics in solution were explored. HD is well-structured in solution, and our study shows that it contains three α-helices. Dynamics study indicates that the N- and C-terminal residues of HD are flexible in solution. HD of human SIX1 exhibits molecular interactions with a short double-strand DNA sequence evidenced by the 1 H-15 N-heteronuclear single quantum correlation (HSQC) and 19 F-NMR experiments. Our current study provides structural information for HD of human SIX1. Further studies indicate that this construct can be utilized to study SIX1 and DNA interactions.

Structural insights into the evolution of the RAG recombinase.

Adaptive immunity in jawed vertebrates relies on the assembly of antigen receptor genes by the recombination activating gene 1 (RAG1)-RAG2 (collectively RAG) recombinase in a reaction known as V(D)J recombination. Extensive biochemical and structural evidence indicates that RAG and V(D)J recombination evolved from the components of a RAG-like (RAGL) transposable element through a process known as transposon molecular domestication. This Review describes recent advances in our understanding of the functional and structural transitions that occurred during RAG evolution. We use the structures of RAG and RAGL enzymes to trace the evolutionary adaptations that yielded a RAG recombinase with exquisitely regulated cleavage activity and a multilayered array of mechanisms to suppress transposition. We describe how changes in modes of DNA binding, alterations in the dynamics of protein-DNA complexes, single amino acid mutations and a modular design likely enabled RAG family enzymes to survive and spread in the genomes of eukaryotes. These advances highlight the insight that can be gained from viewing evolution of vertebrate immunity through the lens of comparative genome analyses coupled with structural biology and biochemistry.

Molecular investigation of the tandem Tudor domain and plant homeodomain histone binding domains of the epigenetic regulator UHRF2.

Ubiquitin-like containing PHD and ring finger (UHRF)1 and UHRF2 are multidomain epigenetic proteins that play a critical role in bridging crosstalk between histone modifications and DNA methylation. Both proteins contain two histone reader domains, called tandem Tudor domain (TTD) and plant homeodomain (PHD), which read the modification status on histone H3 to regulate DNA methylation and gene expression. To shed light on the mechanism of histone binding by UHRF2, we have undergone a detailed molecular investigation with the TTD, PHD and TTD-PHD domains and compared the binding activity to its UHRF1 counterpart. We found that unlike UHRF1 where the PHD is the primary binding contributor, the TTD of UHRF2 has modestly higher affinity toward the H3 tail, while the PHD has a weaker binding interaction. We also demonstrated that like UHRF1, the aromatic amino acids within the TTD are important for binding to H3K9me3 and a conserved aspartic acid within the PHD forms an ionic interaction with R2 of H3. However, while the aromatic amino acids in the TTD of UHRF1 contribute to selectivity, the analogous residues in UHRF2 contribute to both selectivity and affinity. We also discovered that the PHD of UHRF2 contains a distinct asparagine in the H3R2 binding pocket that lowers the binding affinity of the PHD by reducing a potential electrostatic interaction with the H3 tail. Furthermore, we demonstrate the PHD and TTD of UHRF2 cooperate to interact with the H3 tail and that dual domain engagement with the H3 tail relies on specific amino acids. Lastly, our data indicate that the unique stretch region in the TTD of UHRF2 can decrease the melting temperature of the TTD-PHD and represents a disordered region. Thus, these subtle but important mechanistic differences are potential avenues for selectively targeting the histone binding interactions of UHRF1 and UHRF2 with small molecules.

Structural Insight into the Binding of TGIF1 to SIN3A PAH2 Domain through a C-Terminal Amphipathic Helix.

TGIF1 is a transcriptional repressor playing crucial roles in human development and function and is associated with holoprosencephaly and various cancers. TGIF1-directed transcriptional repression of specific genes depends on the recruitment of corepressor SIN3A. However, to date, the exact region of TGIF1 binding to SIN3A was not clear, and the structural basis for the binding was unknown. Here, we demonstrate that TGIF1 utilizes a C-terminal domain (termed as SIN3A-interacting domain, SID) to bind with SIN3A PAH2. The TGIF1 SID adopts a disordered structure at the apo state but forms an amphipathic helix binding into the hydrophobic cleft of SIN3A PAH2 through the nonpolar side at the holo state. Residues F379, L382 and V383 of TGIF1 buried in the hydrophobic core of the complex are critical for the binding. Moreover, homodimerization of TGIF1 through the SID and key residues of F379, L382 and V383 was evidenced, which suggests a dual role of TGIF1 SID and a correlation between dimerization and SIN3A-PAH2 binding. This study provides a structural insight into the binding of TGIF1 with SIN3A, improves the knowledge of the structure-function relationship of TGIF1 and its homologs and will help in recognizing an undiscovered SIN3A-PAH2 binder and developing a peptide inhibitor for cancer treatment.

Conserved Structure and Evolution of DPF Domain of PHF10-The Specific Subunit of PBAF Chromatin Remodeling Complex.

Transcription activation factors and multisubunit coactivator complexes get recruited at specific chromatin sites via protein domains that recognize histone modifications. Single PHDs (plant homeodomains) interact with differentially modified H3 histone tails. Double PHD finger (DPF) domains possess a unique structure different from PHD and are found in six proteins: histone acetyltransferases MOZ and MORF; chromatin remodeling complex BAF (DPF1-3); and chromatin remodeling complex PBAF (PHF10). Among them, PHF10 stands out due to the DPF sequence, structure, and functions. PHF10 is ubiquitously expressed in developing and adult organisms as four isoforms differing in structure (the presence or absence of DPF) and transcription regulation functions. Despite the importance of the DPF domain of PHF10 for transcription activation, its structure remains undetermined. We performed homology modeling of the human PHF10 DPF domain and determined common and distinct features in structure and histone modifications recognition capabilities, which can affect PBAF complex chromatin recruitment. We also traced the evolution of DPF1-3 and PHF10 genes from unicellular to vertebrate organisms. The data reviewed suggest that the DPF domain of PHF10 plays an important role in SWI/SNF-dependent chromatin remodeling during transcription activation.

RNA binding protein RBMS3 is a common EMT effector that modulates triple-negative breast cancer progression via stabilizing PRRX1 mRNA.

The epithelial-to-mesenchymal transition (EMT) has been recognized as a driving force for tumor progression in breast cancer. Recently, our group identified the RNA Binding Motif Single Stranded Interacting Protein 3 (RBMS3) to be significantly associated with an EMT transcriptional program in breast cancer. Additional expression profiling demonstrated that RBMS3 was consistently upregulated by multiple EMT transcription factors and correlated with mesenchymal gene expression in breast cancer cell lines. Functionally, RBMS3 was sufficient to induce EMT in two immortalized mammary epithelial cell lines. In triple-negative breast cancer (TNBC) models, RBMS3 was necessary for maintaining the mesenchymal phenotype and invasion and migration in vitro. Loss of RBMS3 significantly impaired both tumor progression and spontaneous metastasis in vivo. Using a genome-wide approach to interrogate mRNA stability, we found that ectopic expression of RBMS3 upregulates many genes that are resistant to degradation following transcriptional blockade by actinomycin D (ACTD). Specifically, RBMS3 was shown to interact with the mRNA of EMT transcription factor PRRX1 and promote PRRX1 mRNA stability. PRRX1 is required for RBMS3-mediated EMT and is partially sufficient to rescue the effect of RBMS3 knockdown in TNBC cell lines. Together, this study identifies RBMS3 as a novel and common effector of EMT, which could be a promising therapeutic target for TNBC treatment.

The CRL4VPRBP(DCAF1) E3 ubiquitin ligase directs constitutive RAG1 degradation in a non-lymphoid cell line.

The development of B and T lymphocytes critically depends on RAG1/2 endonuclease activity to mediate antigen receptor gene assembly by V(D)J recombination. Although control of RAG1/2 activity through cell cycle- and ubiquitin-dependent degradation of RAG2 has been studied in detail, relatively little is known about mechanisms regulating RAG1 stability. We recently demonstrated that VprBP/DCAF1, a substrate adaptor for the CRL4 E3 ubiquitin ligase complex, is required to maintain physiological levels of RAG1 protein in murine B cells by facilitating RAG1 turnover. Loss of VprBP/DCAF1 in vivo results in elevated RAG1 expression, excessive V(D)J recombination, and immunoglobulin light chain repertoire skewing. Here we show that RAG1 is constitutively degraded when ectopically expressed in a human fibroblast cell line. Consistent with our findings in murine B cells, RAG1 turnover under these conditions is sensitive to loss of VprBP, as well as CRL4 or proteasome inhibition. Further evidence indicates that RAG1 degradation is ubiquitin-dependent and that RAG1 association with the CRL4VPRBP/DCAF1 complex is independent of CUL4 activation status. Taken together, these findings suggest V(D)J recombination co-opts an evolutionarily conserved and constitutively active mechanism to ensure rapid RAG1 turnover to restrain excessive RAG activity.

A native mass spectrometry platform identifies HOP inhibitors that modulate the HSP90-HOP protein-protein interaction.

Herein we describe a native mass spectromery protein-peptide model as a competent surrogate for the HOP-HSP90 protein-protein interaction (PPI), application of which led to the qualititive identification of two new peptides capable of in vitro PPI disruption. This proof of concept study offers a viable alternative for PPI inhibitor screening.

p113 isoform encoded by CUX1 circular RNA drives tumor progression via facilitating ZRF1/BRD4 transactivation.

BACKGROUND: Metabolic reprogramming sustains tumorigenesis and aggressiveness of neuroblastoma (NB), the most common extracranial malignancy in childhood, while underlying mechanisms and therapeutic approaches still remain elusive. METHODS: Circular RNAs (circRNAs) were validated by Sanger sequencing. Co-immunoprecipitation, mass spectrometry, chromatin immunoprecipitation (ChIP) sequencing, and RNA sequencing assays were applied to explore protein interaction and target genes. Gene expression regulation was observed by ChIP, dual-luciferase reporter, real-time quantitative RT-PCR, and western blot assays. Gain- and loss-of-function studies were performed to observe the impacts of circRNA-encoded protein and its partners on the lipid metabolism, mitochondrial activity, growth, invasion, and metastasis of NB cells. RESULTS: A novel 113-amino acid protein (p113) of CUT-like homeobox 1 (CUX1) was identified in NB cells treated by serum deprivation. Further validating studies revealed that nuclear p113 was encoded by circRNA of CUX1, and promoted the lipid metabolic reprogramming, mitochondrial activity, proliferation, invasion, and metastasis of NB cells. Mechanistically, p113 interacted with Zuotin-related factor 1 (ZRF1) and bromodomain protein 4 (BRD4) to form a transcriptional regulatory complex, and mediated the transactivation of ZRF1/BRD4 in upregulating ALDH3A1, NDUFA1, and NDUFAF5 essential for conversion of fatty aldehydes into fatty acids, fatty acid ß-oxidation, and mitochondrial complex I activity. Administration of an inhibitory peptide blocking p113-ZRF1 interaction suppressed the tumorigenesis and aggressiveness of NB cells. In clinical NB cases, high expression of p113, ZRF1, or BRD4 was associated with poor survival of patients. CONCLUSIONS: These results indicate that p113 isoform encoded by CUX1 circular RNA drives tumor progression via facilitating ZRF1/BRD4 transactivation.

The intervening domain is required for DNA-binding and functional identity of plant MADS transcription factors.

The MADS transcription factors (TF) are an ancient eukaryotic protein family. In plants, the family is divided into two main lineages. Here, we demonstrate that DNA binding in both lineages absolutely requires a short amino acid sequence C-terminal to the MADS domain (M domain) called the Intervening domain (I domain) that was previously defined only in type II lineage MADS. Structural elucidation of the MI domains from the floral regulator, SEPALLATA3 (SEP3), shows a conserved fold with the I domain acting to stabilise the M domain. Using the floral organ identity MADS TFs, SEP3, APETALA1 (AP1) and AGAMOUS (AG), domain swapping demonstrate that the I domain alters genome-wide DNA-binding specificity and dimerisation specificity. Introducing AG carrying the I domain of AP1 in the Arabidopsis ap1 mutant resulted in strong complementation and restoration of first and second whorl organs. Taken together, these data demonstrate that the I domain acts as an integral part of the DNA-binding domain and significantly contributes to the functional identity of the MADS TF.

Shaking hands is a homeodomain transcription factor that controls axon outgrowth of central complex neurons in the insect model Tribolium.

Gene regulatory mechanisms that specify subtype identity of central complex (CX) neurons are the subject of intense investigation. The CX is a compartment within the brain common to all insect species and functions as a 'command center' that directs motor actions. It is made up of several thousand neurons, with more than 60 morphologically distinct identities. Accordingly, transcriptional programs must effect the specification of at least as many neuronal subtypes. We demonstrate a role for the transcription factor Shaking hands (Skh) in the specification of embryonic CX neurons in Tribolium. The developmental dynamics of skh expression are characteristic of terminal selectors of subtype identity. In the embryonic brain, skh expression is restricted to a subset of neurons, many of which survive to adulthood and contribute to the mature CX. skh expression is maintained throughout the lifetime in at least some CX neurons. skh knockdown results in axon outgrowth defects, thus preventing the formation of an embryonic CX primordium. The previously unstudied Drosophila skh shows a similar embryonic expression pattern, suggesting that subtype specification of CX neurons may be conserved.