1995

WSU/NSF Summer Engineering Institute

for
Level 6-14 Teachers

ADAPTING AN AUTOMATED BIOANALYTICAL SYSTEM TO A COLLEGE INSTRUCTIONAL LABORATORY ENVIRONMENT

Steven Murov
Chemistry Instructor, Modesto Junior College
in cooperation with
Dr. Bernard Van Wie, Mentor, WSU, and Donna Johnson, High School Teacher, GA

Introduction

The WSU/NSF Summer Engineering Institute has been designed to give grade level 6 -14 teachers from the fields of science and mathematics an experience in chemical engineering and related engineering fields. Because the curriculum of chemical engineering students includes considerable course work in chemistry, chemical engineers generally have a good understanding of the methods and goals of the field of chemistry. On the other hand, despite the fact that it is the chemical engineer who takes the discoveries of chemists and finds practical and economical methods of applying the discoveries, chemists generally have had little or no experience with the field of chemical engineering. Thus chemists do not have a significant understanding of the problems involved with scaling up, continuous or flow systems, economics and the other types of challenges encountered by chemical engineers.

From an instructional perspective, the lack of familiarity of chemists with the field of chemical engineering leads to some unfortunate but at least partially preventable consequences. Through the first and even the second year of college, perspective chemical engineering students take chemistry, physics and math courses as well as general education courses but, except for drafting, little if any course work in the field of engineering. Thus their choice of chemical engineering as a college major is often based on information that may be far from reality. It would be very useful if chemistry, physics and math instructors in high school and college had a good understanding of the goals and methods of engineers and could incorporate some of these concepts into their lectures and laboratories. This would give students at least some basis for a more informed decision regarding a major consistent with their desires and abilities. Perhaps some students would choose engineering who otherwise would have avoided it because of misconceptions. Other students who eventually would have become disillusioned with engineering could be saved several years of course work.

One of the goals of he National Science Foundation funded program at Washington State University is to provide grade level 6 - 14 teachers with an experience that demonstrates the application of engineering principles and approaches. Hopefully the teachers will then be able to pass on the experience to their students. The project directed by Dr. Bernard Van Wie involving the development of "A Flow Injection Analysis Based BIOANALYTICAL System" 1 is ideally suited for providing an engineering experience to science teachers. The system is appropriate because the development of an automated blood analysis system is an engineering problem that involves the application of many concepts that are part of introductory chemistry courses. The analyses performed by the instrument meet and surpass many of the criteria used for the selection of instructional laboratory experiments. In support of this, the following rationales are provided:

  1. The system poses a current and real challenge. The health care industry is commonly confronted with emergency situations where analyses of several chemical species are required in short (<15 min.) periods of time. Even when time is not a primary factor, automated multiple analyses free valuable trained technicians for more people intensive assignments. The development of faster instruments for these analyses will provide quicker treatment improving the chances of complete recovery by patients. The design and development of this instrument involves application of basic chemical engineering principles including small automated flow systems. The design challenges provide students with a good example of the problems that confront chemical engineers.
  2. Parts of the system are easily adaptable to instructional laboratory environments and afford several advantages over experiments that are currently part of the traditional chemistry laboratory.
    1. Many of the current chemistry students are either preengineering or prepharmacy students. Unlike most traditional experiments that are selected primarily because of their relationship to chemical principles, the application to clinical chemistry should provide the relevance that will interest and motivate students. In addition, it is possible to apply the analyses to beverages and this provides an additional relevant and motivating experience.
    2. The experiment currently performed to give students experience with visible spectroscopy usually involves taking the visible absorption spectrum of cobalt(II) and then verifying that its absorption conforms to Beer's Law (A = Ebc). Unfortunately, cobalt(II) is quite toxic and the relevance of this experiment other than as a demonstration of spectroscopy is not obvious. Even more important, the cobalt(II) experiment as with almost all traditional experiments is presented with step by step instructions and tends to be a "cookbook" experience. It is important that a strong effort be made to stimulate thinking by the students and to have them connect the laboratory experiment to the principles that are being studied in the lecture portion of the course. As will be discussed later, the clinical analysis system affords an opportunity for providing more of a research environment to students.
    3. Good kinetics experiments are generally too lengthy and sophisticated for a 3 hour lab period. As a result, the ones performed have many assumptions and approximations. While they serve to demonstrate the principles, the results are certainly not of research quality. It is possible that some of the clinical assays may be adaptable to good 3 hour kinetic experiments.

Background

Glucose Analysis

The automated analytical system will be designed to perform assays of eight routinely measured chemical species at the rate of 480 tests per hour or 60 plasma samples per hour. These will include albumin, total calcium, carbon dioxide, creatinine, glucose, hemoglobin, total protein and urea. For several reasons, the glucose system has been selected as the primary assay for the preparation of a teaching module.

  1. The relevance of glucose and its plasma concentration as an indicator of hypoglycemia and hyperglycemia (and possibly diabetes) are well known to most people.
  2. The basic methods have been worked out by Ian Byrnes2 and Sameer Parab3. In addition, two 1994 Summer Institute students (Don Rogers and Deanne Crichton) worked on the adaptation of this system to high school and middle school laboratories.
  3. The assay involves spectroscopy, solution preparation and dilution and possibly application of a kinetic analysis.

Several methods have been developed for the analysis of glucose4. The Trinder method was selected by Dr. Van Wie's group because it is fast and accurate, relatively easy to use and inexpensive, uses a spectrophotometric analysis technique and is applicable when appropriately used over the range of glucose concentrations desired. Other methods have been suggested for instructional laboratory environments5 but do not appear to be as convenient. The Trinder method also has the advantage that the sample preparation method is completely compatible with the protein analysis method that has been selected. Thus it has been possible to develop an unusual system that includes the analyses of both glucose and protein.

The basic reactions used in the Trinder method are the following:

Trinder Reaction Equations.

A glucose oxidase/peroxidase reagent kit is available from Reagents Applications, Inc., San Diego, CA, that contains 4-aminoantipyrine as the reduced dye. The colorless reduced dye is oxidized in the reaction to a reddish colored dye that has an absorption maximum at 503 nm.

The currently accepted mechanism for the glucose oxidase mechanism was proposed by Duke et al. (1969)6.

Glucose Oxidase Mechanism Reaction Equations

Application of the pseudo-steady-state hypothesis to the above system leads to the following rate expression2:

Rate Equation.

Above temperatures of 30oC, k2 and k5 are reported to be significantly larger than k42,6 thus simplifying the above expression to the following:

Second Rate Equation.

If the first term in the denominator is significantly bigger than the second term, then the expression further simplifies to:

Simplified Rate Equation.

Thus if the assumptions above are valid, the rate should be proportional to the concentrations of the glucose and the glucose oxidase.

Following the production of hydrogen peroxide, peroxidase reacts with the hydrogen peroxide in a two step reaction to produce the oxidized dye that is spectrophotometrically determined at 503 nm.

More Reaction Equations.

Assuming that k6 and k7 are faster than previous steps, the rate expression will remain the same as above.

Glucose analyses are routinely performed as a part of clinical care laboratory tests. Adult plasma glucose concentrations are considered normal in the range 70 - 105 mg/dl (0.0389 - 0.0583 M). Values significantly out of this range can cause health problems and with extreme hyperglycemic conditions (>500 mg/dl) or hypoglycemic conditions (<50 mg/dl), patients are often comatose. There is danger of brain damage especially in hypoglycemic cases. Thus it is very important that the attending physicians obtain a glucose assay in a short time so that treatment can be quickly initiated with insulin for hyperglycemic conditions and with glucose for hypoglycemic conditions.

Analysis of protein.

The analysis for total protein also depends on the addition of a reagent that develops a color with the protein7. The dye, Coomassie Brilliant Blue G-250 (CBB-G) has been selected for use here because of its compatibility with the glucose system and ease of use. The dye has an absorption maximum in the blue region (465 nm) but binds with protein to give an absorption maximum at lower energy (595 nm). The number of dye molecules that bind to each protein is related to the length of the protein chain and the nature of the amino acids present. Thus the proportionality between the protein-dye complex will vary slightly from one protein to another but the measurement still yields a good assay of the total protein present. As 60% of plasma protein is albumin, albumin will be used in this experiment to determine the Beer's law graph.

P + nD = P*nD P = protein

D = dye

n = number of dye molecules that bind to each protein.
This number can vary considerably from one protein to another.

P*nD = protein - dye complex

The measurements are complicated somewhat by the fact that although the peak absorption for the dye is at a considerably higher energy than for the complex, the dye also absorbs at 595 nm. The dye will be used as a blank but its concentration decreases with increasing protein concentration. Using this technique, a graph of A Vs the concentration of protein could deviate from linearity. A linear graph should be obtained if water is used as the blank but absorption readings using this technique are very high and difficult to read on the Spectronic 20.

Approach and Results

Glucose

A. Instrumentation. With the use of a Multistat III (Instrumentation Laboratory), several studies of the glucose - Trinder analysis system were performed to verify the applicability and range of the procedure. In the Multistat system up to 20 samples are loaded into a sample disk with a dilution ratio and reagent amounts programmed before the loader is started. The disk is then transferred to a spectroscopic analyzer that is capable of taking absorption readings every 3 seconds on each of the samples using a preselected filter (for all runs reported here, the filter system had a maximum transmittance at 500 nm). After an initial spinning of the disk where the temperature is increased to the preset temperature (30oC in all of the runs to be reported here), the disk rotation is abruptly stopped causing the reagent to be mixed with the glucose sample. The instrument then resumes the spinning of the sample and takes readings according to the programmed schedule.

For all runs, the loader was programmed to dilute 3 uL of the glucose solution with 24 uL of water. In a second compartment on the disk, the loader added 216 uL of the Trinder's reagent. The analyzer then mixed the two resulting in a 3/243 dilution ratio of the glucose solution.

B. Absorption Spectrum of Oxidized Dye. A solution containing oxidized dye was produced by mixing 30 uL of glucose solution (300 mg/dL), 240 uL of water and 2160 uL of Trinder's reagent. This gives the same 3/243 dilution used in the Multistat. The absorption spectrum was run in a 1 cm cell in a Beckman spectrophotometer.

Absorbtion Graph.

C. Run 1 - Verification of a fit to Beer's Law. A stock solution containing 5000 mg/dL glucose was diluted to 500 mg/dL. This solution was diluted to yield the relative concentrations given in Appendix 1. The graph included shows that a linear Beer's law plot (15 min.values) was obtained through the highest concentration used (300 mg/dL before a 3/243 dilution) with a slope of 35.5 L/mol. This value is identical to the one reported by Byrnes8. Appendix 2 gives absorption vs time plots at two different glucose concentrations. At 30oC, the glucose concentration reaches a relatively flat plateau by 10 minutes.

D. Run 2 - Absorption vs concentration of glucose reagent; End point analyses. For these runs, the prediluted glucose concentration was always 300 mg/dL. The relative glucose reagent concentrations were 1.0, 0.75, 0.50, 0.25 and 0.10. Triplicate samples were studied as a function of time and glucose reagent concentration. In all cases, the 3 samples agreed within experimental error. Appendix 3 gives graphs of the absorption vs time up to 100 minutes for two different glucose reagent concentrations. For the suggested reagent concentration, glucose has a flat plateau between about 20 and 40 minutes but the graphs presented exaggerate changes. The absorption does decrease gradually after the plateau is reached. Probably readings any time in the range of 15 minutes to 60 minutes at 30oC runs will give reliable measurements.

The absorption vs glucose reagent concentration (using 30 minute concentrations) graph (Appendix 4) shows that the maximum absorption falls off sharply when the reagent concentration is decreased to less than 1/2 of its recommended value. Byrnes9 states that there is enough reagent for 875 mg/dL. Since we were using 300 mg/dL, a reagent concentration of 300/875 = 0.34 should be limiting. This is consistent with our observations of a big drop off between 0.5 and 0.25

E. Run 3 - Absorption vs concentration of glucose reagent: Initial rates. The same solutions as above were used again except data was collected every 3 seconds for the first 2 minutes. Examples of absorption vs time at two different reagent concentrations are given in Appendix 5. For relative reagent concentrations of 1.0, 0.75 and 0.50, the graphs were linear over the first 24 seconds. For the two lower concentrations (0.25 and 0.10), an initiation period seemed to occur. "Initial rates" were determined from the linear portion of the graphs after the initiation period. Appendix 6 is a graph of the "initial rates" vs the relative reagent concentration and exhibits a rather smooth curve. However, the kinetics predict a linear dependence on the reagent concentration. This could be due to the fact that the reagent contains several chemicals in addition to the glucose oxidase including buffer, dye and peroxidase. It is possible that the changes in the concentrations of these chemicals is causing the deviation from linearity.

F. Run 4 - Absorption vs glucose concentration: Initial rate studies. This run was designed to test the dependence of the initial rate on a wide range of glucose concentrations. Appendix 7 gives examples of the absorption vs time for three glucose concentrations. Initial rates were obtained by determining the slope of these curves for the first 24 seconds. In all cases, the lines were linear within experimental error over this range. It can be observed however, that for 1000 mg/dL, 24 seconds is the limit of linearity and may be a little beyond it. A graph of the initial rates vs glucose concentration (Appendix 8), yielded a very good straight line with the regression points almost superimposable on the experimental data. Thus this data is consistent with the kinetic expression.

G. Run 5 - Absorption vs glucose concentration: End point analyses. A run with concentrations identical to those in Run 4 was performed but data was collected every 15 minutes for 4 hours. Graphs of the absorption at 60 minutes (Appendix 9) and 4 hours (Appendix 10) vs glucose concentration show a linear relation up to 400 mg/dL and scatter after 500 mg/dL. There are probably two causes of the poor data after 500 mg/dL.

  1. According to Byrnes, the reagent will become limiting at 875 mg/dL. The specification sheet for the reagent (taking into account the dilution factors used) indicates that above 270 mg/dL, glucose samples should be diluted for reliable results.
  2. The solubility of oxygen in water can be calculated from the following equation:10

    ln X1 = -66.73538 + 87.47547/(T/100K) + 24.45264 ln(T/100K) (ref. 10)

    Using the above equation, the mol fractions of oxygen dissolved in water at 1.00 atmosphere of oxygen and 30.0oC and 37.0oC are 2.122x10-5 and 1.934x10-5 respectively. Using densities of water at these two temperatures of 0.99565 g/mL and 0.99333 g/mL and an atmospheric oxygen content of 20.95% yields oxygen concentrations of 2.46x10-4 and 2.23x10-4 mol/L. These numbers do not take into account the fact that some of the air is water vapor and that Pullman has an elevation of about 2300 ft. above sea level. Assuming that the air contains 2% water and that atmospheric pressure in Pullman is 0.93 atm, the values for oxygen concentrations come out 2.24x10-4 and 2.03x10-4 mol/L. These numbers are close to those obtained by Byrnes11 (using an O2 electrode) of 2.06x10-4 and 1.82x10-4 mol/L. For an analysis on the Multistat using a 3/243 dilution factor, 300 mg/dL of glucose corresponds to 2.06x10-4 moles/L. Thus assuming oxygen does not continually enter the system, oxygen becomes limiting above 300 mg/dL. With this in mind, it is surprising that the end point graph (see next paragraph) is linear to 400 mg/dL and it is not surprising that there is scatter above 500 mg/dL.

A graph of the absorption vs glucose concentration up to 400 mg/dL (Appendix 11) did give a very linear relationship and the slope (34.9 L/mol) agreed favorably with the value obtained in Run 1 and the value of Byrnes. The Trinder method (with a 3/243 dilution) can be used with end point analysis up to 400 mg/dL and to 1000 mg/dL with initial rate studies (to 24 seconds). Beyond these conditions, samples should be diluted first.

H. Run 6. Adapting the glucose analysis to the Spectronic 20. The adaptation of the glucose analysis to the Spectronic 20 involves one major consideration. The absorption scale on the instrument is only readable to two significant figures reliably between 0 and about 0.9. Unless the instrument is equipped with a cell holder capable of holding square minicells (path length = 0.5 cm), final glucose concentrations are restricted to a maximum of about 0.14 mM. A crude experiment indicated that the light beam in a 13 mm diameter cuvette goes through the tube between 1.25 and 2.54 cm up from the bottom. This requires a little more than 2 mL to achieve so it was decided to use 3.0 mL samples. Another restriction is that the glucose reagent should be used near its recommended concentration as we have observed that it becomes limiting below 0.5 of its recommended concentration. Therefore, it was decided to use 0.50 mL of the glucose solution and 2.5 mL of the reagent which should keep the reagent concentration at a sufficient level. The most concentrated glucose concentration is to contain 0.14 mM after dilution and using the 1/6 dilution factor results in a glucose concentration of 0.15 g/L. Using a glucose solution with relative dilution factors of 0.2, 0.4, 0.6, 0.8 and 1.0 resulted in the data and graph in Appendix 12. Despite the use of test tubes rather than cuvettes, the regression line is almost superimposable on the experimental points. The method therefore appears to be useable.

A sample of Sunkist Orange was analyzed using the graph in Appendix 12. The sample was diluted 1000x and gave an absorption of 0.74 or 0.12 g/L after dilution or 120 g/L. This is 82% of the value the label indicates for all sugars (35 g/240 mL or 146 g/L). The label also indicates that the source of the sugar is sugar and/or high fructose corn syrup. By definition, high fructose corn syrup contains 58% glucose and 42% fructose. This leads to the conclusion that most of the sugar in this bottle was introduced as glucose. A quick test showed that fructose and sucrose do not give positive glucose readings using Trinder's reagent.

Protein

For protein analysis, the instructions in Reference 7 were directly adaptable. We experienced considerable difficulty preparing a dye solution with an absorption spectrum similar to the ones in the literature7. The literature spectrum shows two peaks with maxima at 465 nm and about 650 nm in a 2:1 intensity ratio. While we always obtain those peaks, their intensities have been variable and the most favorable we could obtain was about 3:2 (the bigger the ratio, the better for analysis of the complex peak at 595 nm). A stock solution containing 0.15 g/L of albumin was prepared and diluted by factors of 0.1, 0.2 ... 1.0. Absorption vs relative protein concentration (Appendix 13 - dye blank and Appendix 14 - water blank) gave close to linear graphs. At higher protein concentrations the data is somewhat scattered and it is possible that the slope is decreasing. This is not unexpected as the dye blank is not the right blank for each point due to dye depletion with increasing protein. Although the water blank system should give a straight line, the data is not easy to read on a Spectronic 20. The minimum absorption is about 0.6 and values above 0.8 because of the logarithmic nature of the readings are all very close together. Despite this difficulty, we did obtain a graph almost identical in appearance to the one with the dye blank.

A preliminary test was performed to see if either glucose or protein interferes with the test of the other and this test was negative.

Glucose + Protein

A pilot experiment with conditions similar to those to be use by students was run. The protein was reduced from the earlier value of 0.15g/L to 0.10g/L to keep absorption values under the reliability limit of about 0.9. A stock solution containing 0.1574 g of glucose and 0.1000 g of protein was diluted to 1 L with buffer solution. Dilution factors of 0.2, 0.4, 0.6, 0.8 and 1.0 were used to prepare solutions for analysis. Solutions of 6 beverages (diluted 1000x) and lactose (0.16 g/L) were also tested. Results are given in Appendices 15 and 16. The beverages were diluted by delivering 5 uL from a pipette into 5 mL of buffer. We had originally intended to use Drummond micopipets for this purpose but repeated weighings of samples using these capillaries was not precise and very inaccurate. An expensive micropipeter was used but these are not generally available and even this instrument was not too precise when tested by weighing water samples. Thus the results below could have some error. Probably the instructor should prepare the diluted beverages by diluting 0.100 mL to 100 mL with buffer. Having the students do this would use too much buffer.

The protein analysis (Appendix 15) again as in the protein test above gave a very good linear plot for absorption vs protein concentration. The beverage results were fairly consistent with the beverage labels. Schweppes and Gatorade did give very small protein readings even though protein is not supposed to be present. Slice and Sunkist gave zero readings as expected. Nonfat milk gave a protein reading 57% of the label value. This could be due to a dilution error or because albumin was used as the standard and other proteins are present in milk. According to Reference 7 (Lott, et.al.), CBB-G gives values that are on the average 63% of biuret results. This is consistent with our results. The Ultra Slim Fast was very "gorpy" and difficult to use. Because of the considerable amount of solid present, it is very difficult to obtain reproducible samples. The only advantage of this sample is that it contains both protein and glucose but its disadvantages outweigh its advantages.

The glucose analysis (Appendix 16) also gave a very good linear plot for absorption vs glucose concentration. The milk and the lactose samples both gave zero amounts of glucose. Apparently under the experimental conditions, lactose does not break down into glucose and galactose and there must not be any significant amounts of free glucose in milk. Slice, Schweppes and Gatorade yielded glucose values of 44%, 33% and 48% of label sugar values. The first two contain "high fructose corn" syrup and/or sugar. Since high fructose corn syrup is 58% glucose and there could be some sucrose present, the values appear to be reasonable. The Gatorade listed sucrose, glucose and fructose and the 48% is certainly possible. The Ultra Slim Fast as explained was difficult to sample so the result is essentially meaningless. The Sunkist orange result (129 g/L vs 146 g/L on the bottle) was consistent (120 g/L) with the result in Run 6 that indicated that much of the sugar was added in the form of glucose (72%) and only 28% as high fructose corn syrup.

Based on maximum absorption readings obtained for glucose and protein of 0.95 and 0.69, it is recommended that the stock solution be changed from 0.15 g of glucose and 0.10 g of protein to 0.12 g of each.


1Grant funded by the Washington Technology Center, Seattle, WA, and Spectrum Systems, Inc., WA

2Ian Byrnes, Kinetic Considerations and Process Limitations for Flow Injection Analysis of Glucose, Washington State Univ. Thesis with Dr. B.J. Van Wie, 1995.

3Sameer Parab, Washington State Univ. Thesis with Dr. B.J. Van Wie, 1995.

4See for example, reference 3 and W. T. Caraway and N. B. Watts, in Fundamentals of Clinical Chemistry, 3rd ed., ed. N. W. Tietz, Saunders, pp. 422-447, (1987).

5a. E. C. Toren, Jr., J. Chem. Ed., 44, 172(1976), b. T. L. Daines, K. W. Morse, J. Chem. Ed., 53, 126(1976).

6Duke, F. R., Weibel, M., Page, D. S., Bulgrin, V. G., and Luthy, J., J. Am. Chem. Soc., 91, 3904(1969).

7For a general discussion of protein concentration health effects and analyses, see G. H. Grant, L. M. Silverman and R. H. Christenson, in Fundamentals of Clinical Chemistry, 3rd ed., ed. N. W. Tietz, Saunders, pp.291 - 331 (1987). For specific information on the use of CBB-G, see J. A. Lott, V. A. Stephan and K. A. Prichard, Jr., Clin. Chem., 29, 1946 (1983) and M. M. Bradford, Analytic Biochemistry, 72, 248, (1976).

8Ref. 2, page 78.

9Ref. 2, page 51.

10Solubility Data Series, "Oxygen and Ozone", Vol. 7, ed. R. Battino, Pergamon Press (1981).

11Ref. 2., page 36.

12To this author, the term high fructose corn syrup is misleading to consumers as the syrup contains less that half fructose.


Appendix 1



Appendix 2



Appendix 3


 


Appendix 4


Appendix 5



Appendix 6



Appendix 7




Appendix 8




Appendix 9


 


Appendix 10



Appendix 11



Appendix 12



Appendix 13



Appendix 14



Appendix 15




Appendix 16




CLINICAL CHEMISTRY: COMBINED SPECTROPHOTOMETRIC ANALYSES OF GLUCOSE AND PROTEIN

Abstract: Glucose and protein assays are two of the analyses commonly performed in routine and emergency health care laboratory procedures. This experiment presents a spectrophotometric method for glucose and protein that is adaptable to high school and college chemistry laboratory environments. The procedure enables both glucose and protein concentrations to be determined from one set of diluted solutions. While the method was adapted from blood plasma techniques, it can be applied to beverages such as milk and sodas. Several independent student investigations are also suggested and encouraged.

Learning Objectives

By performing this experiment, students should gain experience with:

  1. The use of a spectrometer to determine an absorption spectrum and two Beer's law graphs.
  2. Dilutions and dilution calculations.
  3. Use of volumetric glassware.

Comments and Hints

Several special reagents including Glucose Reagent (GR) for glucose, Coomassie Brilliant Blue G-250 (CBB-G) and Bovine serum albumin (BSA) for protein are required for this experiment. GR (also called Trinder's reagent) is available from:

Reagents Applications Inc.
8225 Mercury CT
San Diego, CA 92111-1203
800 438-6100
Sigma Chemical Co.
P.O. Box 14508
St. Louis, MO 63178-9916
800 325-3010

However, it is rather expensive (about $1.60 per student or group). It is possible that local pathology labs and hospitals might be willing to provide outdated samples. CBB-G is available from Sigma and other biochemical sources. CBB-G is available in several different grades. The medium grade and the lowest grade are much less expensive than the best grade and we found the medium grade worked fine in this experiment. We did not test the lowest grade. Also needed are three beverages solutions diluted 500 fold. Recommended are skim milk and two light colored sodas.

While detailed instructions are given for Parts A - D of this experiment, several suggestions are given for more research oriented investigations (Parts E - L). To provide an incentive for students to perform these experiments, it is suggested that the maximum grade for completion without one of the options be 90% and 110% with at least one of the options.

Discussion

A comatose patient is wheeled into the emergency room and the physician begins his diagnosis. What is one of the first requests the physician makes to the nurses? Probably the physician will ask for an analyses of several chemical species including electrolytes and glucose. For the comatose patient, glucose will be especially important as a glucose deficiency or excess can be the cause of the comatose condition. Because quick and proper treatment can maximize the chances of complete recovery, it is very important that facilities are available to provide prompt responses to requested analyses. To accomplish this goal, chemical engineers working with clinical chemists and pathologists have developed fast automated instruments that are capable of performing multiple analyses from a single injection. These instruments are usually large (floor models that are about 5 feet in width) and relatively expensive (>$100,000). For many applications such as in room or on site, it would be more convenient and appropriate to have less expensive table top models.

A joint venture by Dr. Bernard Van Wie of the Chemical Engineering Department at Washington State University and several companies is in progress to develop a small automated multiple blood analysis instrument. The instrument will use a flow injection system that divides the original sample into several parts and adds color producing reagents to each portion of the sample. Each reagent is chosen because of its ability to selectively bind or react with one chemical species (e.g., glucose or protein) in the sample and as a result produce a detectable color. The colored bubble produced flows between a filtered light source and a detector. By comparing the amount of light absorbed by a sample to the amount absorbed by a standard, the concentration of the chemical can be determined.

It has been possible to adapt part of the methodology that will be used in the automated instrument to the instructional laboratory environment. In today's experiment, you will manually perform two of the assays performed by the instrument. However, the experiment has been designed to simulate the automated method as much as possible. You have probably noticed that most chemical analyses require separation before identification. In sharp contrast and consistent with the requirement that the analysis be quick, in this experiment both the glucose and protein assays are performed from the same diluted sample without a separation. Of the common assays performed, glucose is one of the most important and is conveniently adaptable to the instructional laboratory. Of the remaining common assays, protein was selected because: the analyses methods of glucose and protein are compatible, its analysis wavelength is in the visible region and it is possible to analyze both with the same dilution of a stock solution.

As indicated earlier, abnormal glucose concentrations are often a cause of comatose conditions. For normal adults, glucose concentrations are usually in the range 70 - 105 mg/dL. Concentrations below 50 mg/dL (hypoglycemic) and above 400 mg/dL (extreme hyperglycemic) often lead to comatose conditions. Significant hypoglycemic conditions can lead to brain damage and the condition must be treated quickly. Any hyperglycemic value is symptomatic of diabetes and is capable of causing serious health problems in both the short and long term.

While it is sometimes important to analyze for individual proteins such as albumin, (alpha)1-antitrypsin, (alpha)2-macroglobulin or fibrinogen, only total protein will be assayed in this experiment. The total protein concentration of serum obtained from healthy adults is in the range 6 - 8 g/dL or about 100 times more on a mass basis than the amount of glucose present. About 60% of the protein is albumin with the remaining 40% distributed among many other proteins. Abnormally high protein concentrations (hyperproteinemia) can be due to dehydration (e.g., from vomiting), Addison's disease or diabetic acidosis. Low protein concentrations (hypoproteinemia) occur for example in water intoxication or salt retention syndromes or during massive intravenous infusions.

Analysis of glucose.

Several methods have been developed for the analysis of glucose1. The Trinder method was selected because it is fast and accurate, relatively easy to use and inexpensive, uses a spectrophotometric analysis technique and is applicable when appropriately used over the range of glucose concentrations desired. The basic reactions used in the Trinder method are the following:

Trinder Equations.

A glucose oxidase/peroxidase reagent kit is available from Reagents Applications or Sigma Chemical Co., that contains 4-aminoantipyrine as the reduced dye. The colorless reduced dye is oxidized in the reaction to a reddish colored dye that has an absorption maximum in the green region that you will determine.

The currently accepted mechanism for the glucose oxidase mechanism was proposed by Duke et al. (1969)2.

Reaction Equations.
More Reaction Equations.

This rather complicated looking system can be simplified considerably if it is assumed that the first step is significantly slower than all subsequent steps and that there are adequate concentrations of all the reagents including oxygen. The rate expression for product formation under these conditions should be the rate of the first step:

Simplified Rate Equation.

Thus if the assumptions above are valid, the rate should be proportional to the concentrations of glucose and the glucose oxidase.

Again assuming that there are sufficient concentrations of all the reagents, the total amount of oxidized dye produced should be equivalent to the number of moles of glucose initially present. Therefore it should be possible to determine the amount of glucose present either by measuring the amount of oxidized dye produced at the end point (at 30oC the reaction is near completion after 15 min.) or at any specified time during the reaction. Alternatively, the glucose concentration should be proportional to the initial rate of oxidized dye production but this is more difficult to measure.

To determine the concentration of the oxidized dye, you will first determine the absorption spectrum of the oxidized dye. From the spectrum you will select the optimum wavelength for a concentration Vs absorption study. Finally you will measure the absorption of a series of standard solutions of glucose oxidize and prepare a calibration curve (Beer's law graph). At a specific wavelength, the amount of light absorbed is proportional to the concentration of the absorbing species (Beer's law: A = Ebc where A = measured absorption, E = the extinction coefficient of the absorbing material at the selected wavelength, b= the path length of the light through the sample and c = the concentration of the absorbing species). A graph therefore of the absorption Vs the concentration should give a straight line with slope Eb. As you will be determining the absorption in round tubes of the same diameter, b will be a constant. However, its value can only be approximated in round tubes so it is better to simply determine the product of the two constants, E and b, than to try to determine them individually. Once the graph and/or slope is determined, it is possible to determine glucose concentration from absorption measurements of solutions with unknown glucose concentrations.

Analysis of protein.

The analysis for total protein also depends on the addition of a reagent that develops a color with the protein. The dye, Coomassie Brilliant Blue G-250 (CBB-G) has been selected for use here because of its compatibility with the glucose system and ease of use. The dye has an absorption maximum in the blue region (465 nm) but binds with protein to give an absorption maximum at lower energy (595 nm). The number of dye molecules that bind to each protein is related to the length of the protein chain and the nature of the amino acids present. Thus the proportionality between the protein-dye complex will vary slightly from one protein to another but the measurement still yields a good assay of the total protein present. As 60% of plasma protein is albumin, albumin will be used in this experiment to determine the Beer's law graph.

P + nD = P*nD P = protein

D = dye

n = number of dye molecules that bind to each protein.
This number can vary considerably from one protein to another.

P*nD = protein - dye complex

The measurements are complicated somewhat by the fact that although the peak absorption for the dye is at a considerably higher energy than for the complex, the dye also absorbs at 595 nm. The dye will be used as a blank but its concentration decreases with increasing protein concentration. Using this technique, a graph of A Vs the concentration of protein could deviate from linearity. A linear graph should be obtained if water is used as the blank but absorption readings using this technique are very high and difficult to read on the Spectronic 20. Because of this, we recommend the first technique (use the dye as the blank) and in our laboratory over the range of measurements made, the graph was within experimental error of being a straight line.

Procedure

The following solutions will be available in burets in the laboratory:

  1. A stock solution of glucose (0.150 g) + bovine serum albumin (0.100 g) per liter of buffer solution.
  2. Glucose Reagent (GR) (see Comments section)
  3. Buffer Solution (7.65 g NaCl, 1.27 g Na2HPO4, 0.10 g NaH2PO4, 0.21 g KH2PO4 per liter of water).
  4. Coomassie Brilliant Blue G-250 (CBB-G) solution. Prepare by:
    1. add 0.100 g of CBB-G to 50 mL of 95% ethanol and stir to dissolve.
    2. to the solution above (from a), add 100 mL of 85% phosphoric acid and mix.
    3. dilute the solution (from b) to 1.0 liters with water.
    4. gravity filter the final solution (from c) through a Whatman #1 filter.
  5. Three beverages (skim milk and two sodas are recommended) diluted 500 fold with buffer.

The following equipment and glassware will also be needed:

  1. 24 test tubes (minimum volume about 8 mL)
  2. a minimum of 2 Spectronic 20 cuvettes (or with lower precision 13x100 mm test tubes)
  3. 0.30 and 0.50 mL volumetric measuring tools (graduated 1 mL pipets or syringes or volumetric micropipets)
  4. Spectronic 20 or another comparable spectrophotometer

 

A. Preparation of solutions. Number and label eight test tubes 1 - 8. Prepare the first 5 solutions by using the burets to add the amounts in the table below to 5 test tubes. Be sure to mix thoroughly after adding the solutions to the test tubes. Also from burets, add about 3 mL of three different diluted beverages to tubes 6, 7 and 8. These solutions have been prepared by diluting the beverage 500 fold with buffer solution. Be sure to record the beverage label information about sugar and protein concentrations. Solutions 1 - 8 will be used in Parts B, C, D and F of this experiment.

Tube# Stock Solution
(ml)
Buffer Solution
(ml)
Beverage 1
dilution factor
Beverage 2
dilution factor
Beverage 3
dilution factor
1 1 4
2 2 3
3 3 2
4 4 1
5 5 0
6     0.00200
7       0.00200
8         0.00200

B. Determination of the absorption spectrum of the oxidized dye in the glucose system. Transfer 0.50 mL of solution 5 to a cuvette, add 2.5 mL of the glucose reagent (from the buret), stir and allow the solution to sit for about 15 minutes (solution 5g'). In the meantime, warm up and familiarize yourself with the Spectronic 20 instructions.

  1. With the sample compartment empty, set the zero control so that the meter reads 0.0 on the transmittance scale (infinity on the A scale).
  2. Set the wavelength selector to 650 nm.
  3. Prepare a blank by adding 0.50 mL of buffer solution to a cuvette followed by addition of 2.50 mL of glucose reagent. Insert the blank into the Spectronic 20 and set the Transmittance/Absorbance control to a transmittance of 1 (100%) or an absorbance of 0.0.
  4. Insert solution 5g' and read the absorption (after about 15 minutes are up)i
  5. Decrease the wavelength to 625 nm and repeat steps 2, 3 and 4.
  6. Repeat steps 2, 3 and 4 decreasing the wavelength 25 nm each time down to a reading of 400 nm.
  7. Prepare a graph of the absorption spectrum of the oxidized dye by plotting the absorption on the vertical scale (y axis) versus the wavelength on the horizontal scale (x axis).
  8. From the absorption spectrum, determine the optimum wavelength (lambda) for the Beer's law determination.

C. Absorption Vs concentration (Beer's law graph) for glucose system. Transfer 0.50 mL of each solution from Part A to eight, clean, dry, labeled (1g - 8g) test tubes and 0.30 mL of each solution to a second set of clean dry labeled (1p - 8p) test tubes. Set the "p" tubes aside for Part D. Add 2.50 mL of the glucose reagent to each "g" tube. Thoroughly mix each of the tubes and put them aside until after you have performed Part D of this experiment. After you have finished Part D or at least after waiting 15 minutes after preparing the solution, set the Spectronic 20 to the wavelength selected in Part B and set the limits of the meter as in #'s 1 and 3 in Part B. One by one transfer each of the solutions to a clean and dry cuvette, and read the absorption for each of the solutions (1g - 8g). Note that the Spectronic 20 meter settings do not have to be readjusted this time between readings (why not?). Graph the absorption of the solutions Vs the glucose concentrations in moles/L of the prediluted tubes 1 - 5. Determine the slope of the graph and calculate the concentrations of the glucose in each of the diluted beverage samples. Use the dilution factors for the beverages to calculate the concentration of glucose in each beverage and compare these values to values for sugar concentrations on the containers.

D. Absorption Vs concentration graph for the protein system. Add 3.00 mL of CBB-G solution to each "p" tube. Prepare a blank (solution 0p) by adding 0.30 mL of buffer to a cuvette followed by the addition of 3.00 mL of the dye. Thoroughly mix each of the tubes. Set the Spectronic 20 as in Part B #'s 1 - 3 except set the wavelength at 595 nm and set 100% transmittance with solution 0p. After about 5 minutes transfer each of the solutions to a clean, dry cuvette and read its absorption. (Be sure to go back and finish Part C.) Graph the absorption of the solutions Vs the concentrations of protein in moles/L of the prediluted tubes 1 - 5. If the points appear to fall on a straight line determine the slope of the graph and calculate the concentrations of the protein in each of the diluted beverage samples. If the line is not linear, draw a smooth curve through the data and determine the concentrations of the diluted beverages by reading them from the graph. Use the dilution factors for the beverages to calculate the concentration of protein in each beverage and compare these values to values for protein concentrations on the containers.

The following parts are optional but it is strongly recommended that students perform at least one of the options. For these optional experiments, fewer instructions will be given as students should develop their own approaches to solve the problems presented.

E. Additional beverages. Determine the glucose and protein concentrations in other aqueous solutions such as other beverages or urine samples.

F. Kinetic studies of the glucose system. As mentioned in the Discussion section, it is possible to determine the glucose concentration by measuring the oxidized dye concentration when the reaction is virtually finished as done in Part C above or by a determination of the initial rate of reaction. The latter method can be applied without further dilution to a higher range of glucose concentrations but is substantially more time consuming. According to the rate expression, the initial rate is proportional to the initial glucose concentration. Thus the proportionality constant can be determined by measuring a series of initial rates as a function of glucose concentration. Unfortunately, the reaction as performed in Part C has a linear slope only for about 60 seconds down to about 20 seconds depending on the glucose concentration. By the time the solutions are mixed and inserted into the Spectronic 20, a significant amount of time will be lost and it is difficult to acquire enough data to determine initial rates. However, if solutions from tubes 1 - 5 and the glucose reagent are cooled in an ice bath to 0.0oC prior to mixing, the reaction should be slow enough for reasonable data to be obtained. One problem will be that the tubes will begin to warm up as soon as they are placed in the Spectronic 20 but hopefully enough data can be obtained before this introduces significant error.

To 6 cuvettes, add 0.50 mL of each of the solutions from tubes 1 - 6 (do only beverage 1) and cool for several minutes along with 6 test tubes containing 2.5 mL each of the glucose reagent. Zero the instrument as in Part B at the wavelength selected for Part C. When 0.0oC is attained, quickly transfer the glucose reagent from one test tube to the cuvette containing solution 1. Mix, begin timing simultaneously and insert into the instrument. Take readings every 5 seconds for 1 minute. Measure the temperature of the solution. Repeat with the remaining 5 cuvettes.

For each solution, plot the absorption Vs time and determine the initial slope of the graph. Now plot the initial rates as a function of the prediluted glucose concentrations. Is it linear as the equation predicts? Determine the concentration of glucose in beverage 1 and compare the value to the value from Part C.

G. Establishing conditions for the glucose measurements. The detailed instructions presented to you for Parts C and F took into consideration several experimental boundary conditions. Some of these are the readable absorption range on a Spectronic 20, the concentration of the glucose reagent and the concentration of oxygen in water (look at the overall stoichiometry of the reaction and consider the mole ratio of glucose to reagent and oxygen) and the time required for essentially complete reaction at room temperature. Discuss the boundary conditions imposed by at least one of these factors and if possible perform experiments that verify conditions presented in this experiment or establish new ones. (Hints: For the glucose reagent concentration, ask your instructor for the specification sheet that accompanied the glucose reagent. For oxygen, the Handbook of Chemistry and Physics contains an empirical formula for the concentration of oxygen in water in 100% oxygen at atmospheric pressure as a function of temperature.)

H. Interferences in the glucose and protein analyses. Consider and test methods for determining if any chemicals interfere with the glucose and protein analyses. For example, does the presence of protein affect the absorption readings obtained in the glucose system.

I. Absorption spectra of CBB-G and the protein dye complex. The experimental instructions indicated that the absorption maximum of CBB-G occurs at 465 nm and at 595 nm for the complex. Determine the absorption spectra of both but remember to take into account that the dye and the complex are both present during the protein analyses and that the dye concentration is decreased as a result of complex formation.

J. Use of a spreadsheet to analyze, to graph and determine slopes. Enter the data from Parts B, C and D into a spreadsheet such as Excel or Lotus and use the graphing tool in the spreadsheet to graph the data. For parts C and D, use a regression analysis to determine the slope of the best fit line through the data. Compare the results to data obtained manually from graphs. If you performed part F, spreadsheet use should also facilitate its analysis.

K. The health effects of abnormal glucose and protein levels. Write a research paper on the health effects of either (or both) abnormal glucose or protein levels.

L. Paper chromatography of sugars. It is possible to separate and identify sugars using paper chromatography 3. Make up a stock solution containing 1% each of glucose, fructose, sucrose, galactose, lactose and/or any other sugars available in water or ethanol. Spot Whatman #1 filter paper with each of the samples. Insert the paper into one of the solvents listed below. After developing spray the paper with one of the two reagents listed below.

Potential eluting solvents

  1. n-butanol, acetic acid, water (4:1:5)
  2. n-butanol, acetic acid, water (8:1:3)
  3. n-propanol, ethyl acetate, water (7:1:2)
  4. isopropyl alcohol, n-butanol, water (7:1:2)

Potential developing reagents

  1. 1% NaOH in 60% ethanol
  2. 3 g of p-aminobenzoic acid is added gradually to 5.0 mL of hot phosphoric acid until dissolved. Then 300 ml of n-butanol, acetone water (ratio 10:5:2) is added.

    In both cases, the paper is dried and heated for a few minutes at 105oC. For the NaOH spray, the spots will develop overnight without heating. Also for the NaOH, uv light enhances the visibility of the spots.

Test the technique on the stock solution and some beverages.


1See for example, reference 3 and W. T. Caraway and N. B. Watts, in Fundamentals of Clinical Chemistry, 3rd ed, ed. N. W. Tietz, Saunders, pp. 422-447, 1987.

2Duke, F. R., Weibel, M., Page, D. S., Bulgrin, V. G., and Luthy, J., J. Am. Chem. Soc., 91, 3904 (1969).

3a. I. M. Hais and K. Macek, Paper Chromatography: A Comprehensive Treatise, Czech. Acad. (1963), b. L. R. Croft, J. Chem. Ed., 54, 112 (1977), c. I. Smith, Chromatography and Electrophoretic Techniques, Vol. 1, "Chromatography", pp. 310 - 329 (1969), d. A. S. Saini, J. Chromatog., 24, 484 (1966), e. Handbook of Chromatography, eds. G. Zweig and J. Sherma, Chem. Rubber Co., Vol. 1, p. 364 (1972).


Name____________________________________________Date__________Section__________

Prelaboratory Exercises

  1. Determine the normal adult glucose concentration range in molarity. Show your calculations.



  2. Look up and discuss some of the causes and effects of hypoglycemia.



  3. Look up and discuss some of the causes and effects of hyperglycemia.



  4. Look up and describe Addison's disease.



  5. The absorption (A) is defined as -logT where T is the transmittance of the sample or the ratio of the exiting light intensity to the incident light intensity. Explain why it is equivalent to zero the Spectronic 20 at 0.0 on the transmittance scale infinity on the A scale.



  6. Technically dilutions should be made to a total volume rather than the way it is done here by mixing two measured volumes. For example in Part A, tube 3 is prepared by mixing 3.00 mL of stock solution and 2.00 mL of buffer solution. This is done because the volumetric glassware necessary to do the dilution correctly is expensive and generally not available for the instructional laboratory.
    1. Describe a correct technique for the preparing the solutions.


    2. Does the method used here introduce a significant error? Explain your answer.


  7. To determine the protein concentration, the instructions suggest that you plot the absorption Vs the concentration. At the suggested wavelength, 595 nm both the dye and the protein dye complex. For this reason, the dye solution is used as a blank. However, as the protein dye concentration increases, the dye concentration decreases and the blank is really not the right blank.
    1. Explain the last statement.


    2. Should the absorption versus concentration graph necessarily result in a straight line? Explain your answer.


    3. If water had been used as the blank instead of the dye solution, should a graph of absorption versus concentration result in a straight line. Explain your answer. What are some problems with this technique?



Name____________________________________________Date__________Section__________

Results and Conclusions

A. Preparation of solutions.

  1. Concentration of glucose in stock solution (g/L)
  2. Molecular mass of glucose
  3. Concentration of glucose in stock solution (mol/L)
  4. Concentration of protein in stock solution (g/L)
  5. Dilution calculations
    Tube# Relative
    Dilution Factor
    [glut]*
    (mol/L)
    [prot]*
    (g/L)
    1 _______________ _______________ _______________
    2 _______________ _______________ _______________
    3 _______________ _______________ _______________
    4 _______________ _______________ _______________
    5 _______________ _______________ _______________

    *before dilution in cuvettes

  6. Dilution factor for tubes 6, 7 and 8

B. Determination of the absorption spectrum of the oxidized dye in the glucose system.

wavelength (nm) absorption wavelength (nm) absorption
_______________ __________ _______________ __________
_______________ __________ _______________ __________
_______________ __________ _______________ __________
_______________ __________ _______________ __________
_______________ __________ _______________ __________
_______________ __________ _______________ __________
  1. Plot the absorption versus the wavelength and attach the graph to this report.
  2. What is the optimum wavelength for an absorption Vs concentration study?

C. Absorption Vs concentration (Beer's law graph) for glucose system.

Tube# (before dilution in cuvette)
[glut] (mol/L)
(after dilution)
absorption
0g 0.00 0.00
1g _______________ _______________
2g _______________ _______________
3g _______________ _______________
4g _______________ _______________
5g _______________ _______________
6g   _______________
7g   _______________
8g   _______________
  1. Plot the absorption Vs the glucose concentration and attach the graph.
  2. Does the data obey Beer's Law? Explain your answer.


  3. The absorption data could have been graphed Vs the dilution factor or versus the final cuvette concentrations. What difference would this have made in the graph?


  4. Determine the slope of the graph.
  5. For each of the beverages, determine the concentration of the glucose in the tubes before dilution in the cuvette (glut), the calculated concentration (mol/L) in the original beverage (gluc), the concentration of sugar in each beverage from the information on the beverage container (glub) (assume that all the sugar listed on the container is glucose) and the mol % of glucose in the beverage.
Tube # [glut] [gluc] [glub] mol % glu.
6g __________ __________ __________ __________
7g __________ __________ __________ __________
8g __________ __________ __________ __________
  1. Comment on a comparison of your experimental values for glucose concentrations with those calculated from the values on the containers.
    1. If any of your beverages contained "high fructose corn syrup", comment on this label.



    2. Skim milk contains lactose, a disaccharide composed of glucose linked to galactose. Did your test detect glucose in skim milk. What is the significance of this finding?

 

 

 

D. Absorption Vs concentration graph for the protein system.

Tube # (before dilution in cuvette)
[prot] (g/L)
(after dilution)
absorption
0p 0.00 0.00
1p _______________ _______________
2p _______________ _______________
3p _______________ _______________
4p _______________ _______________
5p _______________ _______________
6p   _______________
7p   _______________
8p   _______________
  1. Plot the absorption Vs the protein concentration and attach the graph.
  2. Does the data obey Beer's Law? Explain your answer.



  3. If the graph is linear, determine the slope of the graph.
  4. For each of the beverages, determine the concentration of the protein in the tubes before dilution in the cuvette (prot), the calculated concentration (g/L) in the original beverage (proc) and the concentration of protein in each beverage from the information on the beverage container (prob). (If the line was linear, use the slope to calculate the values. If it was nonlinear, read the concentrations off of the graph.)
    Tube # [prot] [proc] [prob]
    6p _______________ _______________ _______________
    7p _______________ _______________ _______________
    8p _______________ _______________ _______________
  5. Comment on a comparison of your experimental values for protein concentrations with those calculated from the values on the containers. Suggest reasons for differences.






E, F, G, H, I, J, K, L.

For these parts devise, complete and attach your own observation, results and conclusion reports.

Suggest any ways you can think of to improve any part(s) of this experiment.