College of American Pathologists
Printable Version

  Clinical Abstracts





November 2009

Michael Bissell, MD, PhD, MPH

  • Bone mineral density in older adult patients with rheumatoid arthritis
  • Recognizing and identifying apoptotic cells with multifunctional nanospheres
  • Distinguishing acute from chronic and resolved hepatitis C virus infections
  • Molecular typing of Neisseria meningitidis
  • Risk factors for early onset of pseudoxanthoma elasticum
  • Adverse changes to stored red blood cells
  • Proteomics and cytokines in rheumatoid arthritis
  • Fatty liver as a component of metabolic syndrome
  • Association between markers of GFR and markers of inflammation
  • Bone mineral density in older adult patients with rheumatoid arthritis Bone mineral density in older adult patients with rheumatoid arthritis

    Data suggest an association between rheumatoid arthritis and osteoporosis. Possible reasons for reduced bone mineral density (BMD) in rheumatoid arthritis patients include inflammation, reduced physical activity, and use of medications. Inflammatory cytokines may play an important role in the localized bone destruction of rheumatoid arthritis and may accelerate generalized bone loss. Systemic signs of inflammation, presence of rheumatoid factor, and greater disability predicted lower BMD in prior studies. An association between glucocorticoid use and reduced BMD among patients with rheumatoid arthritis has been suggested, but low-dose glucocorticoids may retard disease-related bone loss by suppressing inflammation. Previous studies of osteoporosis in rheumatoid arthritis consist of patients from referral centers or several population-based co-horts. The authors compared BMD among rheumatoid arthritis and nonrheumatoid arthritis subjects in the Third National Health and Nutrition Examination Survey (NHANES III, 1988–1994), a representative U.S. population-based sample. They also examined the asso-ciation of several rheumatoid arthritis-related factors with reduced BMD. The authors selected subjects older than 60 who had rheumatoid arthritis from NHANES III using previously described methods. Femoral neck BMD (FN-BMD) measured by dual-energy x-ray absorptiometry was compared for the rheumatoid arthritis (n=106) and nonrheumatoid arthritis cohorts (n=4,277). Multivariable linear regression models included known risk factors for osteoporosis. Further adjusted analyses compared the BMD among subgroups of patients with rheumatoid arthritis, such as those taking meth-otrexate (MTX), with positive rheumatoid factor, and with elevated C-reactive protein (CRP) levels. The authors found that patients with rheuma-toid arthritis more frequently reported poor health, history of falling, cognitive impairment, early menopause, history of chronic obstructive lung disease, higher total calcium intake, and thiazide use than the nonrheumatoid arthritis subjects (all P<.05). Adjusted FN-BMD was similar between the patients with rheumatoid arthritis (0.71 g/cm2) and nonrheumatoid arthritis subjects (0.72 g/cm2; P=.5). Among patients with rheumatoid ar-thritis, reduced BMD tended to be seen with MTX use (0.60 g/cm2; P=.07), CRP above 1 mg/dL (0.66 g/cm2; P=.09), and positive rheumatoid factor in female patients (0.68 g/cm2; P=.056). However, none of these findings reached statistical significance. The authors concluded that among a U.S. population-based representative sample, FN-BMD was similar in patients with and without rheumatoid arthritis. Several characteristics of patients with rheumatoid arthritis may be associated with reduced BMD.

    Kinjo M, Setoguchi S, Solomon DH. Bone mineral density in older adult patients with rheumatoid arthritis: an analysis of NHANES III. J Rheumatol. 2007;34:1971–1975.

    Correspondence: Dr. M. Kinjo at

    [ Top ]

    Recognizing and identifying apoptotic cells with multifunctional nanospheres Recognizing and identifying apoptotic cells with multifunctional nanospheres

    The ability to rapidly recognize and sort cells is vital to the developing fields of cellular therapy and biotechnology. Being able to separate and analyze cancer cells, T cells, and apoptotic cells associated with many kinds of diseases is becoming increasingly important not only for basic re-search but also for drug screening and development and disease diagnosis. Techniques for recognizing and separating specific target cells from a mixed-cell population are needed for such biomedical analyses and clinical applications. Consequently, fluorescent-magnetic-biotargeting multi-functional nanospheres are likely to find a niche in bioanalysis, biomedicine, and clinical diagnosis. The authors have been developing such multi-functional nanospheres for biomedical applications. To do so, they covalently coupled avidin onto the surfaces of fluorescent-magnetic bifunctional nanospheres to construct fluorescent-magnetic-biotargeting trifunctional nanospheres and analyzed the functionality and specificity of these tri-functional nanospheres for their ability to recognize and isolate apoptotic cells labeled with biotinylated annexin V, which recognizes phosphatidyl-serine exposed on the surfaces of apoptotic cells. The authors found that multifunctional nanospheres can be used in combination with propidium iodide staining of nuclear DNA to identify cells at different phases of the apoptotic process. Furthermore, the authors demonstrated that apoptotic cells induced by exposure to ultraviolet light can be isolated from living cells using a magnet at an efficiency of at least 80 percent. These cells can then be visualized easily with a fluorescence microscope. The authors concluded that fluorescent-magnetic biotargeting trifunctional nanospheres can be a powerful tool for rapidly recognizing, magnetically enriching and sorting, and simultaneously identifying different kinds of cells.

    Song E-Q, Wang G-P, Xie H-Y, et al. Visual recognition and efficient isolation of apoptotic cells with fluorescent-magnetic biotargeting multifunctional nanospheres. Clin Chem. 2007;33:2177–2185.

    Correspondence: Dai-Wen Pang at

    [ Top ]

    xxxxxxxxxxxxxxxxxx Distinguishing acute from chronic and resolved hepatitis C virus infections

    The diagnosis of incident hepatitis C virus infections has relied on a combination of epidemiological risk factor assessment and screening for serum alanine aminotransferase activity and HCV antibody seroconversion. Identifying incident infection is important in the context of treatment with interferon because therapy initiated within three months after infection prevents chronic infection from developing in nearly all cases. Many new infections remain undiagnosed, however, as they are asymptomatic. Furthermore, asymptomatic, primary infections more often tend to be associated with lower rates of spontaneous viral clearance than do symptomatic infections. Therefore, it is clinically important to differentiate incident HCV infections from chronic infections and to identify primary infections that are asymptomatic. Occasional occurrences of acute exacerbation in chronic infection complicate decisions regarding whom to treat. There are no reliable serological or virological assays applicable to acutephase serum samples that can distinguish between acute and chronic HCV infection. A number of studies have shown that serum anti-HCV immunoglobulin M (IgM) is not a useful marker for diagnosing acute infection, unlike with acute hepatitis A and B. The only reliable way to identify incident infections in the laboratory is to demonstrate seroconversion to anti-HCV, which is very difficult to accomplish outside of prospective studies. An alternative approach to the serological diagnosis of primary HCV infections is to measure the avidity of immunoglobulin G (IgG) antibody, since it is well known that the avidity of antibody increases progressively with time after exposure to an immunogen. The authors reported on the development of an anti-HCV IgG avidity test using a novel combination of target antigens and its validation when applied to seroconversion panels and samples taken from chronically infected people and from those with resolved infection. The assay demonstrated an avidity index (AI) value sta-tistically significantly lower in primary HCV infections than in chronic infections. When the assay was applied to past resolved infections, the difference in AI values was not as significant as the difference between incident and chronic infections. The authors concluded that lower AI values obtained in past resolved infections may be more directly related to lower levels of IgG anti-HCV in past resolved infections than in new or chronic infections.

    Klimashevskaya S, Obriadina A, Ulanova T, et al. Distinguishing acute from chronic and resolved hepatitis C virus (HCV) infections by measurement of anti-HCV immunoglobulin G avidity index. J Clin Microbiol. 2007;45:3400–3403.

    Correspondence: Howard A. Fields at

    [ Top ]

    Molecular typing of <em>Neisseria meningitidis</em> Molecular typing of Neisseria meningitidis

    Antigen sequence typing is used widely for highly discriminatory and precise typing of Neisseria meningitidis. Finetyping data are used to monitor epidemiological changes over time and space and to unravel the plethora of circulating antigen types for subcapsular vaccine development. The authors demonstrated the benefits of including FetA antigen typing in the typing regime at the German National Reference Laboratory for Meningococci. For the investigation of infections in 1,616 patients, adding FetA antigen typing increased the number of finetypes 2.3-fold compared to the use of serogrouping in conjunction with PorA typing but without FetA antigen typing. The ferric enterochelin receptor FetA (formerly FrpB) is an immunodominant protein present in outer membrane vesicle vaccines. The authors undertook a study in which they reported the frequency of fetA deletion in Germany, provided evidence by geographic mapping and multilocus ST (MLST) that fetA deletion is a sporadic event, and offered a de-tailed genetic analysis of deletion events. Among invasive meningococcal isolates from 2,201 patients in Germany, the authors identified 11 strains lacking the fetA gene because of deletions mediated by repeat arrays flanking the gene—that is, Correia elements, repeat sequence 13 (RS13), and duplicated RS3. Geographic mapping and multilocus ST of invasive isolates revealed that fetA deletion was a sporadic event without genetic fixa-tion. Among 821 carrier strains, 12 strains lacked fetA, suggesting that it is maintained during asymptomatic carriage. Most of these isolates belonged to the multilocus ST-35 clonal complex (cc). The authors concluded that ST-35 cc strains and ST-192 strains from Burkina Faso, West Africa, may benefit from loss of fetA, but their infrequent occurrence among invasive isolates does not affect fetA antigen ST.

    Claus H, Elias J, Meinhardt C, et al. Deletion of the meningococcal fetA gene used for antigen sequence typing of invasive and commensal isolates from Germany: frequencies and mechanisms. J Clin Microbiol. 2007;45:2960–2964.

    Correspondence: Heike Claus at

    [ Top ]

    Risk factors for early onset of pseudoxanthoma elasticum Risk factors for early onset of pseudoxanthoma elasticum

    Pseudoxanthoma elasticum, or PXE, is an autosomally inherited disorder that affects the skin, eyes, and cardiovascular system. It is characterized by extensive connective tissue alterations, including progressive calcification and fragmentation of elastic fibers and massive accumulation of proteoglycans in the extracellular matrix. PXE, which is also known as Groenblad-Strandberg syndrome, is caused by mutations in the ABCC6 gene. ABCC6 belongs to the ATP-binding cassette transporter subfamily C and encodes a 165-kDa transmembrane protein termed multidrug resistance-associated protein (MRP). To date, more than 100 PXE-associated mutations have been identified in patients of various origins. The vast majority of the mutations are located in the cytoplasmic C-terminal region of the MRP6 protein, close to the second Walker motif that is critical for ATP hy-drolysis. These findings suggest that PXE is caused by impaired MRP6-transport activity. MRPs are involved in hepatic detoxification, drug distribution, and signal transport and carry out the ATP-dependent transport of a variety of compounds. MRP6 production is high in liver, kidney, and, to a lesser extent, tissues affected by PXE. The authors used restriction fragment length polymorphism and allele-specific polymerase chain-reaction analyses to evaluate the distribution of single-nucleotide polymorphisms in the genes encoding catalase (CAT), superoxide dismutase 2 (SOD2), and glutathione peroxidase 1 (GPX1) in DNA samples from 117 German PXE patients and 117 healthy age- and gender-matched controls. The genetic variants investigated had previously been shown to affect the activities of these antioxidant enzymes. The authors found a correlation between genotype and age of disease onset for polymorphisms in CAT (c.–262C>T), SOD2 (c.47C>T), and GPX1 (c.593C>T). Furthermore, age at disease on-set was inversely correlated with the number of mutated alleles, indicating a cumulative effect on the time of disease onset (mean [standard deviation] age, 40.9 [13.6] years, 32.4 [16.3] years, and 25.7 [15.9] years for carriers of zero, one to two, and more than two mutated alleles, respectively; P=.03). The authors concluded that increased oxidative stress due to activity-affecting polymorphisms in genes encoding antioxidant enzymes leads to earlier onset of PXE.

    Zarbock R, Hendig D, Szliska C, et al. Pseudoxanthoma elasticum: genetic variations in antioxidant genes are risk factors for early disease onset. Clin Chem. 2007;53(10):1734–1740.

    Correspondence: Dr. Christian Gotting at

    [ Top ]

    Adverse changes to stored red blood cells Adverse changes to stored red blood cells

    Despite widespread use in the United States, human blood products are licensed only on the basis of the procedures used for collecting, processing, and storing them. Mandated, specific testing ensures safety from infectious diseases as well as compatibility between blood product and recipient, but there is no regulatory or clinical efficacy standard specifying the clinical outcomes constituting an effective transfusion. Furthermore, red blood cell (RBC) transfusions have not been subjected to the formal risk/benefit analysis that is routine for new biological therapeutics. RBCs can be stored up to 42 days under controlled conditions before transfusion. However, they undergo numerous changes during storage, collectively referred to as the “storage lesion,” which may alter their biological function, including delivery of oxygen to cells. Retrospective cohort studies have found a correlation between RBC storage duration and morbidity and mortality rates after transfusion, suggesting progressive storage lesions may be responsible for adverse outcomes. Despite these observational data, no large controlled clinical trials have been conducted to evaluate the relationship between age of stored RBCs and clinical outcomes. A better understanding of the nature and timing of molecular and functional changes in stored RBCs may provide strategies to improve the balance of benefit and risk for RBC transfusion. To this end, the authors analyzed changes occurring during RBC storage, focusing on RBC deformability, RBC-dependent vasoregulatory function, and S-nitrosohemoglobin (SNO-Hb), through which hemoglobin O2 desaturation is coupled to regional increases in blood flow in vivo (hypoxic vasodilation). For the study, 500 mL of blood from each of 15 healthy volunteers was processed into leukofiltered, additive solution three-exposed RBCs and stored between 1°C and 6°C according to AABB standards. The blood was subjected to 26 assays at 0, 3, 8, 24, and 96 hours and at 1, 2, 3, 4, and 6 weeks. RBC SNO-Hb decreased rapidly (1.2 x 10-4 at three hours versus 6.5 x 10-4 [fresh] mol S-nitrosothiol [SNO]/mol Hb tetramer, P=.032, mercuric-displaced photolysis-chemiluminescence assay) and remained low over the 42-day period. The decline was corroborated by using the carbon monoxide-saturated coppercysteine assay (3.0 x 10-5 at three hours versus 9.0 x 10-5 [fresh] mol SNO/mol Hb). In parallel, vasodilation by stored RBCs was significantly depressed. RBC deformability assayed at a physiological shear stress decreased gradually over the 42-day period (P<.001). The authors concluded that time courses vary for several storage-induced defects that might account for observations linking blood transfusions with adverse outcomes. Of clinical concern is that SNO levels and their physiological correlate, RBC-dependent vasodilation, become depressed soon after collection, suggesting that even fresh blood may have developed adverse biological characteristics.

    Bennett-Guerrero E, Veldmant TH, Doctor A, et al. Evolution of adverse changes in stored RBCs. Proc Natl Acad Sci USA. 2007;10(43):17063–17068.

    Correspondence: Timothy J. McMahon at

    [ Top ]

    Proteomics and cytokines in rheumatoid arthritis Proteomics and cytokines in rheumatoid arthritis

    Rheumatoid arthritis involves multiple molecules and pathways. Autoantibodies and cytokines represent classes of immune cell-secreted proteins postulated to play a variety of roles in rheumatoid arthritis, from regulating the initiation and perpetuation of chronic inflammatory responses to joint destruction. However, the precise mechanisms leading to the expression of autoantibodies and cytokines in early rheumatoid arthritis are not completely understood. Although evidence indicating that autoantibodies are directly pathogenic in rheumatoid arthritis is scant, autoantibodies represent important markers for diagnosing and classifying rheumatoid arthritis. By contrast, autoantibodies have been observed infrequently in other types of arthritis. Proinflammatory cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin (IL)1, probably play important roles in regulating immune activation, driving the inflammatory process and promoting joint destruction in a variety of inflammatory joint diseases. Chemokines are chemotactic cytokines produced by fibroblast-like synoviocytes, cells of the innate immune system, and other immunoregulatory cells. Evidence indicates that, among their many roles, they are important potentiators of autoimmune arthritis. Because expression of cyto-kines and chemokines in synovial tissue occurs early in the course of rheumatoid arthritis, they are being evaluated as biomarkers in early rheumatoid arthritis. The advent of proteomics technologies has allowed large-scale analysis of proteins to identify biomarkers and delineate disease subtypes of rheumatoid arthritis and provided insight into the mechanisms underlying these subtypes. The authors developed and applied antigen microar-rays for diagnosing and classifying rheumatoid arthritis and early rheumatoid arthritis. They described 1,536-feature arthritis antigen arrays containing 225 peptides and proteins representing candidate autoantigens in rheumatoid arthritis. Antigens included a wide variety of native and in vitro citrullinated proteins and peptides, which were robotically printed to the surface of microscope slides, where the binding of serum autoanti-bodies was detected. For their study, the authors described a multiplex analysis of serum cytokines using an optimized cytokine bead assay and integration of these datasets with previously determined antigen array-derived autoantibody signatures. They tested hypotheses that cytokines and chemokines derived from subsets of immunoregulatory cells are selectively upregulated in early rheumatoid arthritis and that classes of cytokines are associated with distinct patterns of autoantibody reactivity. They tested serum samples from 56 patients with a diagnosis of rheumatoid arthritis of less than six months’ duration. They also determined cytokine profiles in samples from patients with psoriatic arthritis (PsA) and ankylosing spondylitis (n=21), as well as from healthy people (n=19). They analyzed data using the Kruskal-Wallis test with Dunn’s adjustment for multiple comparisons, linear correlation tests, significance analysis of microarrays, and hierarchical clustering software. The authors found that distinct anti-body profiles were associated with subgroups of patients who exhibited high serum levels of TNF-α, IL1β, IL6, IL13, IL15, and granulocyte macro-phage colony-stimulating factor. Significantly increased autoantibody reactivity against citrullinated epitopes was observed in patients within the cytokine “high” subgroup. Increased levels of TNF-α, IL1α, IL12p40, IL13, and the chemokines eotaxin/CCL11, monocyte chemoattractant protein-1, and interferon-inducible protein 10 were present in early rheumatoid arthritis as compared with controls (P<.001). Chemokines showed some of the most impressive differences. Only IL8/CXCL8 concentrations were higher in patients with PsA/ankylosing spondylitis (P=.02). The authors concluded that increased blood levels of proinflammatory cytokines are associated with autoantibody targeting of citrullinated antigens and sur-rogate markers of dis-ease activity in patients with early rheumatoid arthritis. Proteomic analysis of serum autoantibodies, cytokines, and chemo-kines allows patients with early rheumatoid arthritis to be stratified into molecular subgroups.

    Hueber W, Tomooka BH, Zhao X, et al. Proteomic analysis of secreted proteins in early rheumatoid arthritis: anti-citrulline autoreactivity is associated with upregulation of proin-flammatory cytokines. Ann Rheum Dis. 2007;66:712–719.

    Correspondence: Dr. W. Hueber at

    [ Top ]

    Fatty liver as a component of metabolic syndrome Fatty liver as a component of metabolic syndrome

    Metabolic syndrome represents a cluster of risk factors that are thought to increase the risk of cardiovascular diseases and type 2 diabetes more than the individual components of abdominal obesity, increased serum triglycerides, low high-density lipoprotein (HDL) cholesterol, hyperglycemia, and hypertension. The pathophysiology of metabolic syndrome is believed to include insulin resistance, but it is not clear why only select individuals develop the syndrome. Insulin normally inhibits the production of glucose and very low-density lipoprotein (VLDL) in the liver. Fat accumulation in the liver due to nonalcoholic causes, called nonalcoholic fatty liver disease, is associated with insulin resistance. The ability of insulin to inhibit glucose and VLDL production results in overproduction of glucose and VLDL particles, leading to hypertriglyceridemia and low HDL cholesterol concentrations. The fatty liver also overproduces C-reactive protein and plasminogen activator inhibitor-1, which have been associated with metabolic syndrome. The authors conducted a study in which the components of metabolic syndrome, as defined by the International Diabe-tes Federation, and liver fat content by proton magnetic resonance spectroscopy were measured in 271 nondiabetic subjects (162 women and 109 men). The authors also measured other features of insulin resistance, such as serum insulin and C-peptide; intra-abdominal and sc fat by magnetic resonance imaging; and liver enzymes, including serum alanine aminotransferase and serum aspartate aminotransferase. The authors found that liver fat was four-fold higher in subjects who had metabolic syndrome (n=116; median, 8.2 percent [interquartile range, 3.2–18.7 percent]) than in those who did not (n=155; 2.0 percent [1.0–5.0 percent]; P<.0001). This increase in liver fat remained significant after adjusting for age, gender, and body mass index. All components of metabolic syndrome correlated with liver fat content. The best correlate was waist in women (r=0.59, P<.0001) and men (r=0.56, P<.0001). Liver fat correlated significantly with concentrations of serum alanine aminotransferase (r=0.39, P<.0001 for women; r=0.44, P<.0001 for men) and aspartate aminotransferase (r=0.27, P=.0005 for women; r=0.31, P=.0012 for men). The best correlates of liver fat were fasting serum insulin (r=0.61, P<.0001 for women and men) and C-peptide (r=0.62, P<.0001 for women and men). The authors concluded that liver fat content is significantly increased in people with metabolic syndrome as compared with those without the syndrome, independent of age, gen-der, and body mass index. Of the other markers, serum C-peptide is the strongest correlate of liver fat.

    Kotronen A, Westerbacka J, Bergholm R, et al. Liver fat in the metabolic syndrome. Endocrinol Metab. 2007;92:3490–3497.

    Correspondence: Hannele Yki-Jarvinen at

    [ Top ]

    Association between markers of GFR and markers of inflammation Association between markers of GFR and markers of inflammation

    The use of serum creatinine as a marker of kidney function is limited by factors that influence creatinine concentrations, such as age, gender, race, and weight. While an improvement over serum creatinine in predicting glomerular filtration rate (GFR), formulae such as the Cockcroft-Gault or Modification of Diet in Renal Disease to calculate estimated GFR (eGFR) also have limitations and may be unreliable in certain populations. Cystatin C is a nonglycosylated cysteine protease inhibitor with a molecular weight of 13 kDa that appears to be produced at a constant rate by all nucleated cells. It is freely filtered by the renal glomerulus and metabolized by the proximal tubule. Some studies have suggested that cystatin C may be superior to serum creatinine or creatinine-based estimating equations as a marker of kidney function, but others have found no difference. A study by the Prevention of Renal and Vascular End-Stage Renal Disease (PREVEND) cohort suggested that cystatin C was significantly associated with C-reactive protein (CRP), smoking, and body mass index, even after adjusting for creatinine clearance levels. The authors concluded that cystatin C levels were influenced by these factors, in addition to kidney function. Longitudinal studies have shown that cystatin C has a stronger and more linear association with cardiovascular disease and mortality outcomes compared with creatinine-based measures. These findings led to the hypothesis that the link between cystatin C and inflammation could explain its advantage over creatinine as a prognostic marker. The authors compared the relative strengths of association of cystatin C and eGFR with two inflammatory biomarkers, CRP and fibrinogen, and determined whether these associations were independent of measured creatinine clearance. They measured serum cystatin C and creatinine in 990 outpatients with coronary artery disease enrolled in the Heart and Soul Study. GFR was estimated by the abbreviated Modification of Diet in Renal Disease equation. The authors compared the associations of serum cystatin C and eGFR with CRP and fibrinogen after adjusting for 24-hour creatinine clearance. The authors found that cystatin C concentrations had moderate correlations with CRP (r=0.15; P<.001) and fibrinogen (r=0.26; P<.0001). EGFR had similar correlations with CRP (r=-0.17; P=.01) and fibrinogen (r=-0.25; P<.001) among people with eGFR of 60 mL/min. or less but had no association with either biomarker among those with eGFR of 60 mL/min. or greater (r=0.04; P=.32 and r=-0.03; P=.38). Quartiles of cystatin C were strongly and directly associated with CRP (P=.02) and fibrinogen (P<.007) after multivariate adjustment. However, these associations disap-peared after adjusting for creatinine clearance (P=.26 and .23, respectively). The authors concluded that cystatin C concentrations have moderate associations with CRP and fibrinogen that are not independent of creatinine clearance. Although a gold standard for kidney function is lacking, this analysis suggests that cystatin C captures an association of mildly impaired kidney function with increased inflammation.

    Singh D, Whooley MA, Joachim H, et al. Association of cystatin C and estimated GFR with inflammatory biomarkers: the Heart and Soul Study. Nephrol Dial Transplant. 2007;22:1087–1092.

    Correspondence: Dr. Michael G. Shlipak at

    [ Top ]

    Clinical pathology abstracts editor: Michael Bissell, MD, PhD, MPH, professor, Department of Pathology, Ohio State University, Columbus.