Hemoglobin precipitation greatly improves 4-methylumbelliferone-based diagnostic assays for lysosomal storage diseases in dried blood spots
Abstract
Derivatives of 4-methylumbelliferone (4MU) are favorite substrates for the measurement of lysosomal enzyme activities in a wide variety of cell and tissue specimens. Hydrolysis of these artificial substrates at acidic pH leads to the formation of 4-methylumbelliferone, which is highly fluorescent at a pH above 10.
When used for the assay of enzyme activities in dried blood spots the light emission signal can be very low due to the small sample size so that the patient and control ranges are not widely separated. We have investigated the hypothesis that quenching of the fluorescence by hemoglobin leads to appreciable loss of signal and we show that the precipitation of hemoglobin with trichloroacetic acid prior to the measurement of 4- methylumbelliferone increases the height of the output signal up to eight fold. The modified method provides a clear separation of patients’ and controls’ ranges for ten different lysosomal enzyme assays in dried blood spots, and approaches the conventional leukocyte assays in outcome quality.
1. Introduction
For many years sugar derivatives of 4-methylumbelliferone (4MU) have been favorite substrates for the measurement of lysosomal enzyme activities in a wide variety of cell and tissue specimens. Hydrolysis of these artificial substrates at acidic pH leads to the formation of 4-methylumbelliferone, which is strongly fluorescent at a pH above 10 [1–3]. The optimal excitation wavelength is 365 nm, and the emission is optimal at 448 nm. Thanks to the sensitivity of optical systems and electronic sensors for the light emission signal, 4MU-based substrates are ideally suited for the measurement of lysosomal enzyme activities in samples of small size even down to the level of single cells [4]. This virtue has been successfully employed for the measurement of lysosomal enzyme activities in dried blood spots on Guthrie cards [5–12]. The easiness of sample transport is a clear advantage of the blood spot analysis over other routine diagnostic procedures. In addition, blood spot technology enables high through- put analysis and newborn screening for lysosomal storage diseases [13–19].
Several reports have demonstrated the feasibility of lysosomal enzyme assays in blood spots using 4MU-based substrates for both diagnostic purposes as well as for newborn screening. However, not all assays are fully satisfactory in that the activities of many lysosomal enzymes in blood spots are extremely low so that long incubation times are required and subtle differences between low versus no enzyme activity are lost. In addition, it is not uncommon that the patients’ samples generate lower signals than the substrate blank, which usually is a punch from an empty circle on the Guthrie card [20–23]. In particular cases, the diagnostic outcome can be challenged by poor separation of the patients’ and controls’ ranges [5,9,20–23].
We have investigated the hypothesis that quenching of the fluorescence by hemoglobin leads to appreciable loss of signal, and we have demonstrated that the sensitivity of the procedure is strongly enhanced by precipitation of hemoglobin prior to measuring the 4- methylumbelliferone.
2. Materials and methods
2.1. Blood spot collection
Blood spots were prepared from heparin-blood samples that were sent to our laboratory for diagnostic testing. For the purpose of this study we only used blood spots from patients who were positively diagnosed on the basis of enzyme deficiency in leukocytes or cultured fibroblasts. The patient population included infants, adolescents and adults.The blood was applied with a pipet on Whatman 903 filter paper (Guthrie cards as used for newborn screening) from the middle of the circles till the circles were completely filled (approximately 60 μl).
The spots were dried for at least 17 h (overnight) and the cards were then stored in a sealed plastic bag at −20 °C until use. Blood samples from anonymous healthy individuals were obtained from the Sanquin blood supply foundation in Rotterdam, The Netherlands, and handled in the same way. In each assay we used 30 blood spots from healthy individuals and a variable number (n) of patients who had been previously diagnosed using routine diagnostic procedures. The activities of the following 10 enzymes were determined: α-galacto- sidase A (Fabry disease; n = 6), α-glucosidase (Pompe disease; n = 20), N-acetyl-α-glucosaminidase (Sanfilippo B disease n = 13), α-L-iduronidase (Hurler/Scheie syndrome; n = 12), β-galactosidase (GM1-gangliodsidosis; n = 1), β-hexosaminidase A (Tay–Sachs dis- ease; n =4), β-hexosaminidase A+B (Sandhoff disease; n = 2), chitotriosidase (“lysosomal disease marker”; n = 38), iduronate-2- sulfatase (Hunter disease; n = 8) and palmitoyl-protein thio-esterase 1 (CNL I; n = 4).
Fig. 1. Lysosomal enzyme activities in bloodspots with and with haemoglobin precipitation. The panels A–F show representative experiments in which theactivities of sixdifferent lysosomal enzyme activities were measured in bloodspots from healthy individuals (open circles) and patients with selected lysosomal enzyme deficiencies (closed circles). A: α-galactosidase A activity (AGAL, Fabry disease); B: α-glucosidase activity (AGLU, Pompe disease); C: α-N-acetylglucosaminidase activity (AHEX, Sanfilippo B disease); D: α-L-iduronidase activity (AIDU, Hurler disease); E: iduronate-2-sulfatase activity (IDU2S, Hunter disease); F: palmitoyl-protein thioesterase 1 activity (PPT). Note: in each panel the activities as measured with (right scale) and without (left scale) TCA precipitation of haemoglobin are plotted on a different scale (y-axis).
2.2. Analytical procedures
For each enzyme assay two 3-mm diameter disks were punched from the bloodspot using a paper puncher (McGill Co, Marengo, IL) leaving sufficient clearance from the edge to ensure that the complete punch was saturated with blood. The two punches were placed in a separate 1.5 ml reaction vial, and for all enzyme assays–except for chitotriosidase–20 μl of distilled water and 40 μl of a solution containing the appropriate buffer, the desired 4MU-substrate and other additives if required (see below) were added to it. The vials were incubated for 17 h at 37 °C. The vials were then placed on ice in a pre-cooled aluminum holder, and protein precipitation was per- formed by addition of 20 μl of a 16% (w/v) ice-cold solution of trichloroacetic acid in distilled water. After 10 min the vials were centrifuged for 5 min at 10,000 g at 4 °C, and 60 μl of the supernatant was transferred to a 96-well Optiplate (Perkin Elmer). Finally, 200 μl of 0.5 M sodiumcarbonate/sodiumbicarbonate buffer (pH 10.7) con- taining 0.25% (w/v) Triton X-100 was added to all samples to enhance the fluorescence of the reaction product 4-methylumbelliferone (4MU). For the chitotriosidase assay, 10 μl of distilled water and 100 μl substrate mixture were added to the punches. Following 1 h of incubation at 37 °C the vials were placed on ice and 27.5 μl of a 16% (w/v) ice-cold solution of trichloroacetic acid was added. After 10 min the vials were centrifuged for 5 min at 10,000g at 4 °C and a 110 μl aliquot of the supernatant was transferred to a 96-well Optiplate. Finally, 150 μl of 0.5 M sodiumcarbonate/sodiumbicarbonate buffer (pH 10.7) containing 0.25% (w/v) Triton X-100 was added. The fluorescence was measured with a fluorimeter (Varioskan, Thermo Electron Corporation, Waltham, MA) using an excitation wavelength of 365 nm and an emission wavelength of 448 nm. The activity of each enzyme is expressed in picomoles 4MU per 3 mm diameter disk per 17 h (pmol/punch/17 h). All assays were performed in duplicate and a 4MU standard solution was included.
2.3. Substrates and reaction conditions
For each enzyme assay, the substrate used and a literature reference describing the method is given: α-iduronidase [24]: 4MU- α-L-iduronide (Glycosynth), iduronate-2-sulphatase [25]: 4MU-α- iduronide-2-sulphate (Moscerdam), N-acetyl-α-glucosaminidase [26]: 4MU-acetamido-2-deoxy-α-D-glucopyranoside (Moscerdam), N-acetyl-β-hexosaminidase A [27]: 4MU-6-sulfo-2-acetamido-2- deoxy-β-D-glucopyranoside (Moscerdam), total β-hexosaminidase [28]: 4MU-2-acetamido-2-deoxy-β-D-glucopyranoside (Glycosynth), α-glucosidase [29]: 4MU-α-D-glucopyranoside (Melford), α-galacto- sidase [30]: 4MU-α-D-galactopyranoside (Sigma), β-galactosidase [31]: 4MU-β-D-galactopyranoside (Koch-light), palmitoyl-protein thioesterase [32]: 4MU-6-thiopalmitoyl-β-D-glucoside (Moscerdam), and chitotriosidase [33] 4MU-chitotriose (TRC). 4-Methylumbelliferone (4MU) was purchased from NBS Biologicals).
3. Results and discussion
In search for an improvement of existing methods we noticed that an internal 4MU standard gave a much lower signal when incubated with a punch from a blood-filled circle from the Guthrie card than with a punch from an empty circle. Apparently, the high hemoglobin content of the blood spot extract quenches the fluorescent signal. The problem is particularly prominent when small incubation volumes are being used as in microtiterplate assays. We then investigated whether the signal to noise ratio could be improved by eliminating the hemoglobin via precipitation with ice-cold trichloroacetic acid (TCA). The results of different enzyme assays are shown in Fig. 1 (6 assays; activities expressed in pmoles MU /punch/17 h) and Table 1 (10 assays; raw data in fluorescence units). TCA precipitation of hemoglobin prior to the measurement of 4MU increased the output signals 3 to10-fold depending on the enzyme activity that was measured. In 4 assays (α-galactosidase, α-glucosidase, β-hexosamin- idase A and palmitoyl-protein thioesterase) hemoglobin precipitation eliminated the output of signals lower than the substrate blank, and in 8 assays (α-galactosidase, α-glucosidase, α-glucosaminidase, β- galactosidase, β-hexosaminidase A, total β-hexosaminidase, iduro- nate-2-sulphatase and palmitoyl-protein thioesterase) it improved the separation of the activity ranges of the patients and the healthy individuals. Likewise, TCA precipitation improved the separation between the normal chitotriosidase activities in blood spots of controls and the increased activities in blood spots of patients with lysosomal storage diseases. In our hands, the assay of α-iduronidase activity without hemoglobin precipitation was truly problematic. The emission signal was in many cases far below the substrate blank. Four of the 30 control samples had activities within the patient range (n = 9). With hemoglobin precipitation the signal was raised above the blank and all control samples were clearly separated from the patient samples. TCA precipitation was not absolutely necessary for the assay of β-hexosaminidase A and total β-hexosaminidase as the activity of these enzymes is so high, that the negative effect of quenching does not pose a problem.
The results are summarized in Table 1 reporting for each enzyme assay the raw data, in actual fluorescence units. This way the effect of TCA precipitation is readily demonstrated. A separate column shows the activity ranges that are obtained with conventional assays in leukocytes, also expressed as raw data.
Our results clearly demonstrate that the presence of hemoglobin in the bloodspot extracts greatly interferes with the fluorescence of 4- methylumbelliferone. The sensitivity of blood spot assays employing 4-methylumbelliferyl substrates is substantially improved by TCA precipitation of the hemoglobin prior to the measurement of the 4MU concentration. Thanks to the greatly improved signal to noise ratio, readings lower than the blank are mostly avoided and low activities become measurable. The activities in blood spots of patients remain low, while the activities of healthy individuals are enhanced leading to an appreciable better separation of the activity ranges of patients and controls. The 10-min precipitation step does not lengthen the whole procedure to a great deal as incubation times of up to 17 h are commonly used to obtain reliable readings. Notably, when the results of our improved blood spot assays are compared with those of conventional leukocyte assays, the actual readings in terms of fluorescence units obtained after 1-h incubation for the conventional assay or 17 h for the blood spot assay are very much the same.
4. Conclusion
We recommend the TCA precipitation step for all enzyme assays on blood spots in which 4-methylumbelliferyl derivatives are used as artificial substrate because it substantially enhances the fluorescence output signal and improves the separation of the controls’ and patients’ activity ranges. The blood spot assay and the leukocyte assay now approach each other in outcome quality. Our new method seems applicable for both routine diagnostic purposes as well as for newborn screening.