The molecular electronic device and the biochip computer: present status.

The idea that a single molecule might function as a self-contained electronic device has been of interest for some time. However, a fully integrated version--the biochip or the biocomputer, in which both production and assembly of molecular electronic components is achieved through biotechnology-is a relatively new concept that is currently attracting attention both within the scientific community and among the general public. In the present article we draw together some of the approaches being considered for the construction of such devices and delineate the revolutionary nature of the current proposals for molecular electronic devices (MEDs) and biochip computers (BCCs). With the silicon semiconductor conductor industry already in place and in view of the continuing successes of the lithographic process it seems appropriate to ask why the highly speculative MED or BCC has engendered such interest. In some respects the answer is paradigmatic as much as it is real. It is perhaps best stated as the promise of the realm of the molecular. Thus it is envisioned that devices will be constructed by assembly of individual molecular electronic components into arrays, thereby engineering from small upward rather than large downward as do current lithographic techniques. An important corollary of the construction technique is that the functional elements of such an array would be individual molecules rather than macroscopic ensembles. These two aspects of the MED/BCC--assembly of molecular arrays and individually accessible functional molecular units--are truly revolutionary. Both require scientific breakthroughs and the necessary principles, quite apart from the technology, remain essentially unknown. It is concluded that the advent of the MED/BCC still lies well before us. The twin criteria of utilization of individual molecules as functional elements and the assembly of such elements remains as elusive as ever. Biology engineers structures on the molecular scale but biomolecules do not seem to be imbued with useful electronic properties. Molecular beam epitaxy and thin-film techniques produce electronic devices but they "engineer down" and are currently unable to generate individual molecular units. The potential of the MED/BCC field is matched only by the obstacles that must be surmounted for its realization.

Binding of photobilirubin to human serum albumin. Estimate of the affinity constant.

Results of experiments based upon circular dichroic spectra suggest that configurationally (Z leads to E) isomerized bilirubin (photobilirubin) binds to human serum albumin at the primary bilirubin binding site with an affinity only 2-3 times lower than that of bilirubin. The high affinity of photobilirubin for albumin, comparable to that of bilirubin, supports the roles of albumin in the stabilization and transport of the isomerized pigment in vivo and strongly suggests that albumin also functions to sequester photobilirubin effectively, reducing its toxic potential. The high affinity of photobilirubin for albumin predicts that the isomerized pigment, formed in large amounts during phototherapy for neonatal hyperbilirubinemia, should not appear in the fast-diazo-reacting ('direct') bilirubin pool nor should it interfere with nonspectroscopic bilirubin binding tests. These predictions were confirmed for the Evelyn and Malloy diazo assay for 'direct' bilirubin and a Sephadex chromatography method for assessing 'loosely bound' plasma bilirubin.

Relationships of bilirubin binding parameters.

The apparent unbound bilirubin concentration by the "peroxidase" method (U) and the total unconjugated bilirubin in blood (T), albumin-bound bilirubin (B), and reserve bilirubin binding capacity (R) by the bilirubin hematofluorometer were measured in 164 specimens from 98 neonates and in a series of artifactual specimens, made by adding bilirubin to the blood of a single adult donor. Linear correlations between U and (B/R) were found for both the prepared specimens (r = 0.99) and the clinical specimens (r = 0.87), with the slopes of the regression lines being close to the reciprocal of the albumin-bilirubin binding constant, a prediction of the mass action law. An excellent linear correlation was observed for the prepared specimens (r = 0.96) between U and (T--B), the concentration of bilirubin bound to low-affinity secondary sites ("loosely bound bilirubin"). A simple model for low-affinity binding of bilirubin in blood predicts this simple relation. A significant linear correlation between U and (T--B) was found for the clinical specimens, although the correlation was less good (r = 0.72), as one would expect. The demonstrated simple linear relationships between U, (B/R), and (T--B) support the hypothesis that both the hematofluorometer and peroxidase methods provide valid measurements of bilirubin binding status.

Fluorescence studies of protoporphyrin. Transport and clearance.

By means of spectrofluorometry porphyrins can be assayed directly in body fluids without extraction. Rapid diagnosis of various primary and secondary porphyrinemias can be made based upon such fluorometric measurements. Fluorometric methods can also provide information about porphyrin binding sites that is useful in understanding porphyrin transport and clearance. Free erythrocyte protoporphyrin produced in the circulating red cells of lead intoxicated Mallard ducks rapidly diffuses from those cells.

Fluorescence methods in the diagnosis and management of diseases of tetrapyrrole metabolism.

Under appropriate conditions fluorescent porphyrins and bilirubin present in blood and other body fluids can be examined spectrofluorometrically without prior extraction. Uses of such direct fluorescence spectroscopy of porphyrins and bilirubin in studies and diagnoses of diseases associated with abnormal or impaired heme synthesis and metabolism are reviewed. The method of "front-face" fluorometry which allows quantitative assays of fluorescent porphyrins and bilirubin in small undiluted blood specimens is described.

Fluorimetric study of the binding of protoporphyrin to haemopexin and albumin.

Fluorescence spectra of protoporphyrin bound to its most affinitive site on human serum albumin, bound to human haemopexin and dissolved in human plasma reveal that, when present in plasma, at least 90% of this porphyrin is bound to albumin. Human serum albumin binds protoporphyrin with an affinity KA = 3 X 10(9)M-1 in phosphate-buffered saline. The affinity of haemopexin for protoporphyrin is 4 times smaller. From these data it is concluded that less than 1% of plasma protoporphyrin is bound to haemopexin. Implications of the data for protoporphyrin transport and clearance are discussed.

Picosecond primary photoprocesses of bilirubin bound to human serum albumin.

We measure the fluorescence quantum yield of bilirubin bound to its highest-affinity site on human serum albumin to increase from about 0.001 near room temperature to 0.5 at 77 K. The quantum yield for configurational (Z leads to E) photoisomerization about the meso double bonds concomitantly decreases from about 0.22 to less than 0.01 over the same temperature range in reciprocal relationship to the fluorescence yield. Transient absorption spectra recorded after excitation with a 0.5-ps pulse of 305-nm light decay with a lifetime of 19 +/- 3 ps at 22 degrees C and 35 +/- 7 ps at 2 degrees C. Bilirubin undergoes the same photoisomerization reaction in chloroform solution, in which a similar short-lived (17 +/- 3 ps at 22 degrees C) transient is observed. From these and other data we conclude that configurational isomerization of bilirubin is the predominant nonradiative pathway that competes with pigment fluorescence, that photoisomerization proceeds via a short-lived (much less than 18 ps) partially twisted excited-singlet-state intermediate, and that bilirubin remains relatively unihibited with respect to photoisomerization when bound to human serum albumin.

Photoisomerized bilirubin in blood from infants receiving phototherapy.

A pigment different from (Z,Z)bilirubin-IX alpha was detected by fluorometric methods in blood specimens from newborn infants undergoing blue-light therapy for unconjugated hyperbilirubinemia; it was not detected in specimens from infants not under therapy. The phototherapy-associated pigment has fluorescence, solubility, and photochemical properties that are identical to those exhibited by what are thought to be configurational (Z leads to E) isomers of bilirubin. It is concluded that isomerized bilirubin in the blood of neonates under phototherapy can reach as high as 15% of the total.

Rapid fluorometric assay of bilirubin and bilirubin binding capacity in blood of jaundiced neonates: comparisons with other methods.

The concentrations of total blood bilirubin, albumin-bound bilirubin, and the reserve and total bilirubin binding capacities of 35 neonatal blood samples (28 patients) were determined by automated front-face fluorometry ((hematofluorometer). These values were compared to results of diazo determinations, Sephadex gel filtration, and peroxidase-oxidation methods. Total blood bilirubin level by fluorometry agreed well with the total plasma bilirubin level by diazotization (r = .96, sigma = 1.7 mg/100 ml). Albumin-bound bilirubin concentrations by fluorometry also correlated well with diazo values (r = .95, sigma = 1.9 mg/100 ml) and were slightly lower than the total blood bilirubin concentrations. Values for total bilirubin binding capacity determined by fluorometry agreed well with results obtained for the same specimens by Sephadex gel filtration (n = 28, r = .97, sigma = 1.8 mg/100 ml) and by peroxidase-catalyzed oxidation (n = 25, r = .97, sigma = 1.7 mg/100 ml). The agreement among the results obtained by the three methods indicates a well-defined in vitro end point at which available primary or "tight" binding sites on albumin are saturated with bilirubin. In this clinical experience the coefficient of variation of results with the hematofluorometer was 8.4% for total blood bilirubin and 6.5% for total binding capacity. A comparison of "sick" with "well" infants revealed that the fraction of bilirubin not bound to albumin was significantly different for these two groups. The assays made with the hematofluorometer are quick (10 to 15 minutes) and require only a small quantity (approximately 150 microliters) of blood.

Hepatic protoporphyrin production in human protoporphyria. Effects of intravenous hematin and analysis of erythrocyte protoporphyrin distribution.

Human protoporphyria with atypical features suggesting increased hepatic protoporphyrin synthesis was investigated in 2 patients. Analysis of the distribution of protoporphyrin among circulating erythrocytes of increasing age indicated that the erythrocyte porphyrin burden derived predominantly from erythroid sources in case 1, and from hepatic sources in case 2. Intravenous hematin was administered to both patients to assess any negative feedback effect on protoporphyrin synthesis. Erythrocyte, fecal and plasma porphyrin levels were measured serially during basal, treatment, and follow-up periods. In case 1, a significant (P less than 0.001) drop in both fecal and plasma levels accompanied hematin, while erythrocyte levels remained unchanged. Hematin produced no appreciable changes in porphyrin concentrations in case 2. Allergic vasculitis followed hematin use in both cases. Based on data of this study and on previous data, a model for protoporphyrin transport and clearance was developed in which the variable clinical and biochemical manifestations of human protoporphyria are related to the relative contributions of erythroid and hepatic sources to the abnormal protoporphyrin pools.

A rapid fluorometric method for determining bilirubin levels and binding in the blood of neonates: comparisons with a diazo method and with 2-(4'-hydroxybenzene)azobenzoic acid dye binding.

A simple, rapid fluorometric method for determining the albumin-bound bilirubin concentration, total blood bilirubin concentration, and the bilirubin reserve-binding capacity of albumin was clinically evaluated using blood specimens from 79 neonates. This study showed that these bilirubin determinations, made by means of the Bell Laboratories hematofluorometer, correlated well with plasma bilirubin levels obtained by a diazotization (Jendrassik-Grof) method. Hematofluorometer reserve-binding capacities correlated very well with 2-(4'-hydroxybenzene)azobenzoic acid (HABA) dye reserve-binding capacities for specimens of artificially jaundiced adult blood. For specimens of neonatal blood the HABA dye reserve capacity was, on the average, higher than the hematofluorometer reserve-binding capacity, particularly for specimens from low-birth-weight babies (less than 2,000 gm). Comparison of HABA reserve capacity and hematofluorometer reserve capacity for high-birth-weight babies (greater than 2,000 gm) gave data very similar to those for adult blood specimens. The specific bilirubin-binding capacity of albumin was found to be greater for infants whose birth weight exceeded 2,000 gm than for the lower birth weight group. The total blood bilirubin concentration obtained by the hematofluorometer is shown to be significantly higher than the concentration of bilirubin bound to albumin, an indication of other important compartments of bilirubin in blood.

Hemoglobin determined in 15 microL of whole blood by "front-face" fluorometry.

We describe the rapid, inexpensive fluorometry of hemoglobin in undiluted whole blood. The procedure consists only of adding a standard quantity of a fluorescent dye to a measured volume (approximately 15 microL) of blood, together with some solubilizing detergent. The assay is based on the attenuation of the dye's fluorescence (excited within the region 400--440 nm) that results from the competitive absorption of exciting light by the hemoglobin present--the "inner filter effect." The wavelength that one can use is optional and will determine which dyes can be used. Measurements are made with the hematofluorometer, a "front-face" filter fluorometer (Blumberg et al., J. Lab. Clin. Med. 89: 712--723, 1977). We demonstrate the validity of the method for two dyes, rhodamine B and fluorescein dibutyrate, which we used with hematofluorometers that were designed to determine blood zinc protoporphyrin and bilirubin, respectively. Our method exhibited a standard error of about 4 g of hemoglobin per liter vs the comparison method (Coulter Counter method, for which the CV is 1.2%). The CV is about 3%. The method seems appropriate for "field" use (i.e., use outside the laboratory) in anemia-screening programs.