Deep-learning reconstruction of complex dynamical networks from incomplete data

Reconstructing complex networks and predicting the dynamics are particularly challenging in real-world applications because the available information and data are incomplete. We develop a unified collaborative deep-learning framework consisting of three modules: network inference, state estimation, and dynamical learning. The complete network structure is first inferred and the states of the unobserved nodes are estimated, based on which the dynamical learning module is activated to determine the dynamical evolution rules. An alternating parameter updating strategy is deployed to improve the inference and prediction accuracy. Our framework outperforms baseline methods for synthetic and empirical networks hosting a variety of dynamical processes. A reciprocity emerges between network inference and dynamical prediction: better inference of network structure improves the accuracy of dynamical prediction, and vice versa. We demonstrate the superior performance of our framework on an influenza dataset consisting of 37 US States and a PM2.5 dataset covering 184 cities in China.

Model-free tracking control of complex dynamical trajectories with machine learning

Nonlinear tracking control enabling a dynamical system to track a desired trajectory is fundamental to robotics, serving a wide range of civil and defense applications. In control engineering, designing tracking control requires complete knowledge of the system model and equations. We develop a model-free, machine-learning framework to control a two-arm robotic manipulator using only partially observed states, where the controller is realized by reservoir computing. Stochastic input is exploited for training, which consists of the observed partial state vector as the first and its immediate future as the second component so that the neural machine regards the latter as the future state of the former. In the testing (deployment) phase, the immediate-future component is replaced by the desired observational vector from the reference trajectory. We demonstrate the effectiveness of the control framework using a variety of periodic and chaotic signals, and establish its robustness against measurement noise, disturbances, and uncertainties.

Effects of growth feedback on gene circuits: A dynamical understanding

The successful integration of engineered gene circuits into host cells remains a significant challenge in synthetic biology due to circuit-host interactions, such as growth feedback, where the circuit influences cell growth and vice versa. Understanding the dynamics of circuit failures and identifying topologies resilient to growth feedback are crucial for both fundamental and applied research. Utilizing transcriptional regulation circuits with adaptation as a paradigm, we systematically study 435 distinct topological structures and uncover six categories of failures. Three dynamical mechanisms of circuit failures are identified: continuous deformation of the response curve, strengthened or induced oscillations, and sudden switching to coexisting attractors. Our extensive computations also uncover a scaling law between a circuit robustness measure and the strength of growth feedback. Despite the negative effects of growth feedback on the majority of circuit topologies, we identify a few circuits that maintain optimal performance as designed, a feature important for applications.

Short-lived chimera states

In the classic Kuramoto system of coupled two-dimensional rotators, chimera states characterized by the coexistence of synchronous and asynchronous groups of oscillators are long-lived because the average lifetime of these states increases exponentially with the system size. Recently, it was discovered that, when the rotators in the Kuramoto model are three-dimensional, the chimera states become short-lived in the sense that their lifetime scales with only the logarithm of the dimension-augmenting perturbation. We introduce transverse-stability analysis to understand the short-lived chimera states. In particular, on the unit sphere representing three-dimensional (3D) rotations, the long-lived chimera states in the classic Kuramoto system occur on the equator, to which latitudinal perturbations that make the rotations 3D are transverse. We demonstrate that the largest transverse Lyapunov exponent calculated with respect to these long-lived chimera states is typically positive, making them short-lived. The transverse-stability analysis turns the previous numerical scaling law of the transient lifetime into an exact formula: the "free" proportional constant in the original scaling law can now be precisely determined in terms of the largest transverse Lyapunov exponent. Our analysis reinforces the speculation that in physical systems, chimera states can be short-lived as they are vulnerable to any perturbations that have a component transverse to the invariant subspace in which they live.

Clinical observation of Porcine Collagen Membrane + artificial Bovine Bone Granules Guided tissue regeneration combined with Autologous CGF in the treatment of severe periodontitis bone defect

Objective

To investigate and analyze the clinical observation of porcine collagen membrane + artificial bovine bone granules guided tissue regeneration (GTR) combined with autologous concentration of growth factors (CGF) in the treatment of severe periodontitis bone defect.

Methods

A total of 94 patients with severe periodontitis bone defects admitted to Shanxi Bethune Hospital from January 2019 to January 2022 were included. They were divided into two groups by simple randomization method. Patients in the control group were treated with porcine collagen membrane + artificial bovine bone granules GTR, while those in the observation group were treated with autologous CGF on the basis of the control group. Before and after treatment, periodontal clinical indicators [sulcus bleeding index (SBI), gingival retreat index (GR), probing depth (PD), clinical attachment loss (CAL), alveolar bone height (AH)] and bone resorption markers [Osteoprotegerin (OPG), bone gla protein (BGP), Type-1 collagen N-terminal peptide (NTX)] were compared between the two groups, and the incidence of postoperative complications in the two groups was recorded.

Results

The total efficacy of observation group was significantly higher than that of control group (p<0.05). Three months after surgery, the observation group had lower levels of SBI, PD, CAL and NTX while higher levels of GR, AH, OPG and BGP than the control group (p<0.05). There was no significant difference in complication rate between the two groups (p>0.05).

Conclusion

Porcine collagen membrane + artificial bovine bone granules GTR combined with autologous CGF boasts various benefits in the treatment of severe periodontitis bone defects, such as improvement of clinical outcomes, amelioration of periodontal tissue and inhibition of bone resorption.

Total removal of a large esophageal schwannoma by submucosal tunneling endoscopic resection: A case report and review of literature

BACKGROUND

Primary schwannoma is a rare submucosal tumor of the esophagus, which is most often benign, and surgery is the only effective treatment. So far, only a few cases have been reported. Herein, we reported a single case diagnosed with primary esophageal schwannoma that was totally removed by submucosal tunneling endoscopic resection (STER).

CASE SUMMARY

A 62-year-old man presented to the hospital with a history of resection of a malignant gastric tumor and mild dysphagia. Endoscopic examination revealed a large submucosal elevated lesion in the esophagus 25-30 cm from the incisors. Endoscopic ultrasonography detected a 45 mm × 35 mm × 31 mm hypoechoic lesion; chest computed tomography showed a mass of approximately 55 mm × 35 mm × 29 mm. A preliminary examination showed features suggestive of a stromal tumor. Pathological findings indicated esophageal schwannoma. Next, STER alone was performed to completely resect the mass, and the patient recovered well post-surgery. Afterward, the patient was discharged and showed no tumor recurrence at 33 mo of follow-up.

CONCLUSION

Endoscopic resection is still an effective treatment for large esophageal schwannomas (> 30 mm) under meticulous morphological evaluation.

Reservoir computing as digital twins for nonlinear dynamical systems

We articulate the design imperatives for machine learning based digital twins for nonlinear dynamical systems, which can be used to monitor the "health" of the system and anticipate future collapse. The fundamental requirement for digital twins of nonlinear dynamical systems is dynamical evolution: the digital twin must be able to evolve its dynamical state at the present time to the next time step without further state input-a requirement that reservoir computing naturally meets. We conduct extensive tests using prototypical systems from optics, ecology, and climate, where the respective specific examples are a chaotic CO₂ laser system, a model of phytoplankton subject to seasonality, and the Lorenz-96 climate network. We demonstrate that, with a single or parallel reservoir computer, the digital twins are capable of a variety of challenging forecasting and monitoring tasks. Our digital twin has the following capabilities: (1) extrapolating the dynamics of the target system to predict how it may respond to a changing dynamical environment, e.g., a driving signal that it has never experienced before, (2) making continual forecasting and monitoring with sparse real-time updates under non-stationary external driving, (3) inferring hidden variables in the target system and accurately reproducing/predicting their dynamical evolution, (4) adapting to external driving of different waveform, and (5) extrapolating the global bifurcation behaviors to network systems of different sizes. These features make our digital twins appealing in applications, such as monitoring the health of critical systems and forecasting their potential collapse induced by environmental changes or perturbations. Such systems can be an infrastructure, an ecosystem, or a regional climate system.

Predicting amplitude death with machine learning

In nonlinear dynamics, a parameter drift can lead to a sudden and complete cessation of the oscillations of the state variables-the phenomenon of amplitude death. The underlying bifurcation is one at which the system settles into a steady state from chaotic or regular oscillations. As the normal functioning of many physical, biological, and physiological systems hinges on oscillations, amplitude death is undesired. To predict amplitude death in advance of its occurrence based solely on oscillatory time series collected while the system still functions normally is a challenge problem. We exploit machine learning to meet this challenge. In particular, we develop the scheme of "parameter-aware" reservoir computing, where training is conducted for a small number of bifurcation parameter values in the oscillatory regime to enable prediction upon a parameter drift into the regime of amplitude death. We demonstrate successful prediction of amplitude death for three prototypical dynamical systems in which the transition to death is preceded by either chaotic or regular oscillations. Because of the completely data-driven nature of the prediction framework, potential applications to real-world systems can be anticipated.

Synchronization within synchronization: transients and intermittency in ecological networks

Transients are fundamental to ecological systems with significant implications to management, conservation and biological control. We uncover a type of transient synchronization behavior in spatial ecological networks whose local dynamics are of the chaotic, predator–prey type. In the parameter regime where there is phase synchronization among all the patches, complete synchronization (i.e. synchronization in both phase and amplitude) can arise in certain pairs of patches as determined by the network symmetry—henceforth the phenomenon of  ‘synchronization within synchronization.’ Distinct patterns of complete synchronization coexist but, due to intrinsic instability or noise, each pattern is a transient and there is random, intermittent switching among the patterns in the course of time evolution. The probability distribution of the transient time is found to follow an algebraic scaling law with a divergent average transient lifetime. Based on symmetry considerations, we develop a stability analysis to understand these phenomena. The general principle of symmetry can also be exploited to explain previously discovered, counterintuitive synchronization behaviors in ecological networks.

Intratumoral Susceptibility Signals Reflect Biomarker Status in Gliomas

Susceptibility-weighted imaging (SWI) can be a useful tool to depict vascular structures in brain tumors as well as micro-bleedings, which represent tumor invasion to blood vessels and could also be representative of tumoral angiogenesis. In this study, we investigated the relationship between SWI features and glioma grades, and the expression of key molecular markers isocitrate dehydrogenase 1 (IDH1), O-6-methylguanine-DNA methyltransferase (MGMT), and 1p19q. The gliomas were graded according to the intratumoral susceptibility signals (ITSS). We used the Mann-Whitney test to analyze the relationship between ITSS grades and the pathological level and status of these markers. Additionally, the area under the curve (AUC) was used to determine the predictive value of glioma SWI characteristics for the molecular marker status. In these cases, the ITSS grades of low-grade gliomas (LGG) were significantly lower than those of high-grade gliomas (HGG). Similarly, the ITSS grades of gliomas with IDH1 mutations and MGMT methylation were significantly lower than those of gliomas with Wild-type IDH1 and unmethylated MGMT. However, ITSS grades showed no relationship with 1p19q deletion status, while they did show significant predictive ability for glioma grade, IDH1 mutation, and MGMT methylation. These findings indicate an association between some molecular markers and cerebral microbleeds in gliomas, providing a new avenue for non-invasive prediction of molecular genetics in gliomas and an important basis for preoperative personalized surgical treatment based on molecular pathology.

Preoperative inflammation markers and IDH mutation status predict glioblastoma patient survival

Recent studies suggest that inflammation response biomarkers are prognostic indicators of solid tumor outcomes. Here, we quantify the prognostic value of the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and lymphocyte-to-monocyte ratio (LMR) in glioblastomas (GBMs), taking into consideration the role of the isocitrate dehydrogenase (IDH) mutation status. We examined 141 primary glioblastomas (pGBMs) and 25 secondary glioblastomas (sGBMs). NLRs, PLRs, and LMRs were calculated before surgery. IDH mutations were detected immunohistochemically after tumor resection, and patients' clinical outcomes were analyzed after classification into GBM, pGBM, and IDH-wild type glioblastoma (IDH-wt GBM) groups. To make comparisons, we set cutoffs for NLR, PLR and LMR of 4.0, 175.0, and 3.7, respectively. In a multivariate analysis, both NLR (HR=1.712, 95% CI 1.026-2.858, p=0.040) and PLR (HR=2.051, 95% CI 1.288-3.267, p=0.002) had independent prognostic value. While a low NLR was associated with a better prognosis only in the IDH-wt GBM group, PLR was predictive of patient survival in the GBM, pGBM, and IDH-wt GBM groups. By contrast, LMR exhibited no prognostic value for any of the 3 types of GBM.

Percolation on networks with weak and heterogeneous dependency

In real networks, the dependency between nodes is ubiquitous; however, the dependency is not always complete and homogeneous. In this paper, we propose a percolation model with weak and heterogeneous dependency; i.e., dependency strengths could be different between different nodes. We find that the heterogeneous dependency strength will make the system more robust, and for various distributions of dependency strengths both continuous and discontinuous percolation transitions can be found. For Erdős-Rényi networks, we prove that the crossing point of the continuous and discontinuous percolation transitions is dependent on the first five moments of the dependency strength distribution. This indicates that the discontinuous percolation transition on networks with dependency is determined not only by the dependency strength but also by its distribution. Furthermore, in the area of the continuous percolation transition, we also find that the critical point depends on the first and second moments of the dependency strength distribution. To validate the theoretical analysis, cases with two different dependency strengths and Gaussian distribution of dependency strengths are presented as examples.

[Beijing common green tree leaves' accumulation capacity for heavy metals]

Seasonal variation of heavy metal contents in leaves and their relationships with soil heavy metal pollution levels were studied through measuring and analyzing the leaves of the common tree species in Beijing and soil heavy metal contents, to detect heavy metal accumulation ability of plant leaves. The results showed that: (1) the contents of Cu, Pb, Zn in plant leaves first decreased and then increased, again declined with changing the seasons (from spring to winter). Cr concentration showed the trend of first increase and then decrease from spring to winter, and the highest in the autumn; the accumulation capacities of Cu for Babylonica and Japonica were higher in the spring, summer and autumn, while Tabuliformis was in winter; the higher accumulation capacities for Cr, Pb were Japonica and Platycladus, and in winter were Platycladus and Bungeana; the higher accumulation capacities for Zn were Babylonica and Bungeana, while Platycladus in winter; (2) the pollution degree of four kinds of heavy metals (Cu, Cr, Pb, Zn) from downtown to suburbs showed that: Jingshan (C =2.48, C is contamination factor) > Olympic (C = 1.27) > Songshan (C = 1.20) > Shuiguan (C = 1. 18); (3) the heavy metals concentration of same plant leaves in the water of the Great Wall changed larger, but those in the other three areas showed that: Jingshan > Olympic > Songshan; the ability of same species leaf to absorb different sorts of heavy metals showed that: Zn >Cu >Pb >Cr; the difference between Zn content and Cr content was significant (P <0.01); (4) the relationship between heavy metal content in plant leaves and soil heavy metal pollution levels presented a quadratic polynomial relation; the significant correlation was found between other three heavy metal contents of plant samples and soil samples, but they were not the case for the Cu, and the correlation coefficients were above 0. 9.

Carboxylic acid functionalized ortho-linked oxacalix[2]benzene[2]pyrazine: synthesis, structure, hydrogen bond and metal directed self-assembly

Cyclooligomerization of 2,6-dichloropyrazine 4 and benzyl 2,3-dihydroxybenzoate 5 under microwave irradiation resulted in a racemic pair of ester functionalized ortho-linked oxacalix[2]benzene[2]pyrazine 6, which was further transformed to the corresponding racemic carboxylic acid functionalized ortho-linked oxacalix[2]benzene[2]pyrazine 3. Both enantiomers of 3 adopt 1,3-alternate conformations with their two carboxylic acid groups pointing to opposite directions in the solid state. Enantiomers of 3 form a step-like one-dimensional supramolecular polymer via intermolecular hydrogen bond interactions between the carboxylic acids for crystals obtained in methanol. No hydrogen bonds were formed between the carboxylic acids for crystals of 3 obtained in pyridine and aqueous guanidine solutions; instead, intermolecular hydrogen bonds between the carboxylic acid groups of 3 and pyridine, as well as guanidinium ions were formed. Under metal-mediated self-assembly conditions, the pyrazinyl nitrogen atoms in 3 interacted with transition metal ions, such as Ag(I), Cu(II) and Zn(II), and resulted in the formation of four new metal-containing supramolecular complexes. Metallomacrocycles 7, 8 and 9 were formed by reactions of 3 with Ag(I) or Cu(II) ions by bridging two ligands 3 in the equatorial region via M-N coordination bonds. A one-dimensional coordination polymer 10 was generated by reaction between ligand 3 and Zn(II) ions, and a cage-based structure is presented in 10 by bridging of the cyclophane units by Zn(2+) ions via Zn-N and Zn-O bonds.

Transcriptional regulation of acetylcholinesterase-associated collagen ColQ in fast- and slow-twitch muscle fibers

The presence of a collagenous protein (ColQ) characterizes the collagen-tailed forms of acetylcholinesterase (AChE) and butyrylcholinesterase at vertebrate neuromuscular junctions, which is tethered in the synaptic basal lamina. ColQ subunits, differing mostly by their signal sequences, are encoded by transcripts ColQ-1 and ColQ-1a, which are differentially expressed in slow- and fast-twitch muscles in mammals, respectively. Both ColQ transcripts are derived from a single COLQ gene. Transcripts encoding ColQ increased during myogenic differentiation of C2C12 cells; the increase was in parallel with AChE catalytic subunit. Quantitative PCR analysis indicated that the increase during the myotube formation was due to the up regulation of ColQ-1 transcript instead of ColQ-1a. In order to reveal the regulatory mechanism of ColQ transcripts, two distinct promoters, pColQ-1 and pColQ-1a, were isolated from human COLQ gene. The ColQ promoters showed a muscle fiber type-specific expression pattern, and which was in line with the expression of endogenous transcript. After in vivo DNA transfection, pColQ-1 showed strong activity in slow-twitch muscle (e.g. soleus), while pColQ-1a was preferably expressed in fast-twitch muscle (e.g. tibialis). Mutation analysis of the ColQ promoters suggested that the muscle fiber type-specific expression pattern of ColQ transcripts was regulated by a slow upsteam regulatory element (SURE) and a fast intronic regulatory element (FIRE). These results explain the specific expression patterns of collagen-tailed AChE in slow and fast muscle fibers.

Identification and characterization of a rat novel gene RSEP4 expressed specifically in central nervous system

The low-abundantly expressed genes composed the majorities of the mRNAs expressed in the central nervous system (CNS), and were thought to be important for the normal brain functions. Through differential screening a low-abundance cDNA sublibrary with mRNA from neuropathic pain of chronic constriction injury (CCI) model, we have identified a novel rat gene, rat spinal-cord expression protein 4 gene (RSEP4). The total length of RSEP4 cDNA is 2006 bp, with a 501 nucleotide open reading frame (ORF) that encodes a 167 amino acid polypeptide. Northern blot revealed that RSEP4 was expressed specifically in the CNS. In situ hybridization showed that the mRNA of RSEP4 was strongly expressed in the CA1, CA2, CA3 and DG regions of hippocampus, the Purkinje cells of cerebellum, and the small sensory neurons of dorsal horn and large motor neurons of ventral horn of spinal cord. Over-expression of RSEP4-EGFP fusion protein in the human embryonic kidney 293T cells showed that RSEP4 protein was mainly localized in the cell cytoplasm. These results suggest that RSEP4 may play some roles in the CNS.

Maternal zinc deficiency impairs brain nestin expression in prenatal and postnatal mice

Effects of maternal dietary zinc deficiency on prenatal and postnatal brain development were investigated in ICR strain mice. From d 1 of pregnancy (E0) until postnatal d 20 (P20), maternal mice were fed experimental diets that contained 1 mg Zn/kg/day (severe zinc deficient, SZD), 5 mg Zn/kg/day (marginal zinc deficient, MZD), 30 mg Zn/kg/day (zinc adequately supplied, ZA) or 100 mg Zn/kg/day (zinc supplemented, ZS and pair-fed, PF). Brains of offspring from these dietary groups were examined at various developmental stages for expression of nestin, an intermediate filament protein found in neural stem cells and young neurons. Immunocytochemistry showed nestin expression in neural tube 10.5 d post citrus (dpc) as well as in the cerebral cortex and neural tube from 10.5 dpc to postnatal d 10 (P10). Nestin immunoreactivities in both brain and neural tube of those zinc-supplemented control groups (ZA, ZS, PF) were stronger than those in zinc-deficient groups (SZD and MZD). Western blot analysis confirmed that nestin levels in pooled brain extracts from each of the zinc-supplemented groups (ZA, ZS, PF) were much higher than those from the zinc-deficient groups (SZD and MZD) from 10.5 dpc to P10. Immunostaining and Western blots showed no detectable nestin in any of the experimental and control group brains after P20. These observations of an association between maternal zinc deficiency and decreased nestin protein levels in brains of offspring suggest that zinc deficiency suppresses development of neural stem cells, an effect which may lead to neuroanatomical and behavioral abnormalities in adults.

Evidence for a mouse brain-specific variant of alpha-tubulin

While studying the neural precursor cell intermediate filament protein known as nestin in the developing mouse brain, we observed a strong cross-reaction of our nestin antibody with a 50 kDa protein that appeared on embryonic day 10 and continued to accumulate until postnatal day 1. Here we report evidence that this protein is a brain-specific variant form of alpha-tubulin and discuss its implications.

[An antibody recognizing neuron specific tubulin]

A 20-residue peptide corresponding to the C-terminal amino acid sequence of rat nestin was synthesized by the solid phase method. The anti-peptide antibody (designated Anti-Nes-2) against nestin was prepared. Western blots showed that Anti-Nes-2 recognized not only mouse nestin with a MW of 240 kD but also a band with a MW of 50 kD. N-terminal amino acid sequence showed that this 50 kD protein is alpha-tubulin. Western blots with Anti-Nes-2 and with monoclonal antibodies against alpha- and beta-tubulin revealed that this 50 kD band could only be detected in different stages of mouse brain and in the primary culture of neural precursor cells (NPCs), with higher expression during the development of mouse brain and the maturation of NPCs; whereas alpha- and beta-tubulin were expressed in different cell lines and tissues of adult mouse. Taken together, these results indicate that 50 kD protein recognized by Anti-Nes-2 is a neuron-specific alpha-tubulin and could be a neuron-specific posttranslational modification isotype of alpha-tubulin.

Reduction of thyroxine uptake into cerebrospinal fluid and rat brain by hexachlorobenzene and pentachlorophenol

In the present study the effects of hexachlorobenzene (HCB) and the metabolite pentachlorophenol (PCP) were investigated with respect to uptake of thyroxine (T4) into cerebrospinal fluid (CSF) and brain structures of rats. [125I]T4 was taken up into CSF of control rats by a relatively slow process, reaching a steady state after about 3 h. Both repeated dosing of HCB and single doses of PCP caused decreased uptake of [125I]T4 into CSF, total brain tissue as well as specific brain structures, such as occipital cortex, thalamus, and hippocampus. Although HCB-treatment caused a build-up of HCB and PCP levels in serum in brain only HCB was present in significant amounts (16% of the serum level). In CSF, both HCB and PCP concentrations were below detection levels. Separate experiments with PCP showed, however, a dose- and time-dependent uptake of PCP into CSF. The present results indicate that PCP and the parent compound HCB are able to affect brain supply of T4. This may have consequences for an adequate development of the brain or proper brain function in adults. The exact mechanisms of interference of PCP and/or HCB in brain uptake of T4 remain to be established.

Differentiation-dependent expression of retinoid-binding proteins in BFC-1 beta adipocytes

Recently, we demonstrated that adipose tissue plays an important role in retinol storage and retinol-binding protein (RBP) synthesis. Our data suggested that RBP expression in adipose tissue is dependent on the state of adipocyte differentiation. To examine this possibility, we explored the differentiation-dependent expression of RBP using BFC-1 beta preadipocytes, which can be stimulated to undergo adipose differentiation. Total RNA was isolated from undifferentiated (preadipocytes) and differentiated (adipocytes) BFC-1 beta cells and analyzed by Northern blotting. RBP mRNA was not detected in the preadipocytes, but considerable RBP mRNA was present in differentiated BFC-1 beta cells. In BFC-1 beta cells, induced to differentiate with insulin and thyroid hormone, RBP mRNA was first detected after 4 days, reached a maximum level by day 10, and remained at this maximum level for at least 2 more days. Cellular retinol-binding protein was expressed at low levels in the BFC-1 beta preadipocytes and the level of expression increased for 6 days after induction to differentiate and slowly declined on later days. Neither the maximum level of RBP expression nor the day on which this level was reached was influenced by the level of retinol provided in the BFC-1 beta culture medium. BFC-1 beta cells secreted newly synthesized RBP into the culture medium at a rate of 43 +/- 14 ng RBP/24 h/10(6) adipocytes. When the BFC-1 beta adipocytes were provided 1.0 microM retinol in the medium, they accumulated the retinol and synthesized retinyl esters. These studies with BFC-1 beta cells confirm that RBP synthesis and secretion and retinol accumulation are intrinsic properties of differentiated adipocytes. Furthermore, they suggest that RBP and cellular retinol-binding protein gene expression are regulated as part of a package of genes which are modulated during adipocyte differentiation.