1. ABSTRACT

The Holder method is the recommended pasteurization method for human milk banks, as it ensures the microbiological safety of human milk (HM). The loss of some biologically active milk components, due to the heat treatment, is a main limit to the diffusion of donor HM. High-temperature short-time (HTST) pasteurization may be an alternative to maintain the nutritional and immunological quality of HM. The aim of the present study was to compare the impact of Holder and HTST pasteurization on the HM protein profile. The protein patterns of HTST-treated milk and raw milk were similar. The Holder method modified bile salt-stimulated lipase, lactoferrin and components of the immune system. The HTST method preserved the integrity of bile salt-stimulated lipase, lactoferrin and, to some extent, of IgAs. Holder pasteurization decreased the amount of bile salt-stimulated lipase and inactivated the remaining molecules, while the HTST method did not alter its activity. Pasteurization increased the bioavailable lysine quantity. HTST pasteurization seems to better retain the protein profile and some of the key active components of donor HM.

2. INTRODUCTION

Increasing evidence shows that the confirmed short- and long-term benefits of human milk (HM) for term infants extend also to preterm and very low birth weight infants (VLBWIs) (1-3). Fresh mother's own milk is the best choice for feeding all newborns, including preterm and VLBWI. Pasteurized donor HM offers a safe alternative and it is recommended as the feeding of choice when mother's own milk is not available or when the quantity is not sufficient (4-5). In many countries, HM banking is well established and standardized, and donor HM has become a health care component for preterm infants (6-9).

Pasteurization ensures the microbiological safety of donor HM, and the Holder method (62.5° C, 30 min) is the recommended method in for use in HM banks (6-9). However, pasteurization inactivates some of the nutritive and biologically active compounds of HM (10-14), which are considered to be responsible for its clinical benefits. This leads to concern that pasteurization may reduce the nutritional quality of HM and its protective effects against infections and necrotizing enterocolitis (NEC) (15), although this has not been observed clinically. Pasteurized donor HM feeding has been shown to reduce feeding intolerance, and to confer protection against NEC and long-term cardiovascular disease risk (16-18). Recently, a pasteurization method, based on heating at a higher temperature but for a shorter time (HTST or Flash pasteurization), and which is in use in the food industry, has been suggested as an alternative to the Holder method for HM pasteurization (19-21).

Among the biologically active components, HM proteins play a key role, as they have nutritive, enzymatic, trophic and immunomodulatory functions (22). During heat treatments, milk proteins can be oxidatively modified, both via the Maillard reaction, which mainly gives lactosylation on lysine residues, and via lipid oxidation, where HNE adducts (4-hydroxy-2-nonenal) can react with histidine, lysine, and cysteine residues. Reactive dicarbonyl compounds or α-β unsaturated aldehydes, originating from these two modification pathways, can lead to protein carbonylation. The occurrence of such modified residues may lead to decreased protein digestibility and reduced nutritional supply (23,24). In previous studies, the Holder method was found to induce carbonylation on amino acid residues (23,24), whereas only one investigation has been conducted to study the effects of the HTST method on the HM protein characteristics (20).

The aim of the present study was to compare the impact of two different pasteurization methods, Holder and HTST, on the protein profile of donor HM.

3. MATERIALS AND METHODS

3.1. Samples

HM samples were collected simultaneously at the Bufalini Hospital (Cesena) from four healthy donor mothers, 1-3 wk after term delivery. The milk specimens were obtained by completely emptying one breast using an electric pump. Immediately after collection, the samples were pooled and divided into 3 aliquots. One aliquot was stored at 4�C until other the two aliquots underwent two different heat treatments. One aliquot was pasteurized by the Holder method (62.5�C for 30 min) (Metallarredinox, Italy) and the other one by the HTST method (72�C for 15 s) (Lauda Ecoline R.E., Brinkmann Instruments, Canada). After these treatments, the three aliquots (from here onwards called Raw, Holder and Flash) were immediately frozen at -80�C for subsequent analyses. Part of each sample was skimmed by centrifugation at 2,000 g for 30 min at 4�C, in order to be used for SDS-PAGE and subsequent immunostaining analyses, while the remaining part of each sample was used for available lysine determination and lipase assay. Quantification of the total protein content on each sample was performed using 2D Quant kit (GE Healthcare).

3.2. Monodimensional electrophoresis (SDS-PAGE)

In order to perform homogeneous SDS-PAGE on the skimmed breast milk, 5 μg of the total proteins for each well were mixed with Laemmli sample buffer (28), without any reducing agent. SDS-PAGE was carried out on 12% polyacrilamide gel (8 x 8.5 cm), with Laemmli running buffer on a BioRad MiniProtean III Cell, at 125 V.

In order to conduct the gradient SDS-PAGE, one well, 4-12% gradient Nupage Novex Bis-Tris (Bis (2-hydroxyethyl) imino-tris (hydroxymethyl) methane-HCl) precast gels (Invitrogen) were used to separate skimmed breast milk samples (360 �g total proteins for each sample), with MES (2-(N-morpholino) ethane sulfonic acid) (Invitrogen) as running buffer, on a Novex Mini-cell (Invitrogen) at 200 V. The gels were stained with a freshly home-made prepared Blue Coomassie Colloidal stain (29) and digitized with a GS-800 Densitometer (Bio-Rad).

3.3. Standard oxidation curves

In order to evaluate the degree of carbonylation of the HM proteins before and after pasteurization, a specific immunostaining was performed. A compound, 2,4-Dinitrophenylhydrazine (DNPH), that specifically binds to carbonyls was used to derivatize the blotted membranes obtained from the homogeneous SDS-PAGE of raw and heat treated milk. A 0-40% gradient of oxidized BSA over fully reduced BSA was used as the reference compound (standard curve). The standard curves were built by mixing variable amounts of HOCl-oxidized BSA (Fluka) with fully reduced BSA, while maintaining a constant total protein concentration, in order to create a 0-40% gradient of oxidized BSA over reduced BSA.

3.3.1. Reduced BSA

Commercially available BSA contains carbonyls that can be reduced by sodium borohydride (NaBH4). Fully reduced BSA has been prepared by reacting a 10 g/L PBS solution (pH 7.4) containing 140 mmol/L of NaCl with 20 g/L NaBH4 for 40 min, followed by neutralization at pH 7.2 with 1 mol/L HCl. After overnight dialysis against 10 mmol/L PBS (pH 7.2), the protein concentration was brought to 4 g/L (30)

3.3.2. Oxidized BSA

105 mg of BSA were dissolved in 15 mL of 10 mmol/L PBS (pH 7.4) containing 0.1 mol/L NaCl, and incubated for 1 h with 7 g/L of sodium hypochlorite (NaOCl) at 37�C, under argon. The fully oxidized BSA solution was passed over a PD10 column (GE Healthcare) to remove unreacted NaOCl, and eluted with PBS to reach a final protein concentration of 4 g/L (31).

3.4. Western Blot on polyvinylidene fluoride (PVDF) membrane

PVDF membranes (Invitrogen) were soaked in methanol for 30 s, then washed with distilled water, and left for few minutes before Western blotting in 0.5X Towbin transfer buffer (32) added with 5% methanol. After homogeneous SDS-PAGE, the proteins were transferred at 30 V from the gel to the PVDF membranes using a tank transfer Xcell II Blot Module (Invitrogen), with 0.5X Towbin transfer buffer added with 5% methanol. The membranes used for the immunochemical analysis were soaked in methanol for 15 s before derivatization, while the negative control membranes were air-dried at room temperature.

3.5. DNP derivatization on membranes

The methanol activated membranes were equilibrated for 5 min in 20% methanol-80% PBS (pH 7.2), and then incubated in 2 mol/L HCl for 5 min, prior to protein derivatization. The membranes were then incubated for 5 min in a 0.5 mmol/L solution of DNPH in 2 mol/L HCl, and washed three times in 2 mol/L HCl for 5 min and seven times in methanol for 5 min each (33,34). Continuous shaking was used during all the incubation and washing steps. The membranes were finally dried overnight at room temperature.

3.6. Hydrophobic immunostaining

The sample and negative control membranes were immunostained using the protocol described by Mansfield (1995) (35), with minor modifications. The derivatized PVDF membranes were incubated for 1 h at 25�C with the primary antibody solution, which consisted of a 1:10,000 dilution of an anti-DNP-KLH rabbit IgG fraction antibody (Molecular Probes) in an incubation solution containing PBS, 0.05% Tween-20 and 1% gelatine (Bio-Rad). the membranes were the washed three times for 2 min with a washing solution containing PBS and 0.05% Tween-20. They were then incubated with a 1:2,500 dilution of horseradish peroxidise (HRP)-Goat Anti-Rabbit IgG (H+L) Conjugate (Invitrogen) in the incubation solution, for 30 min at 25�C. The membranes were washed 3 times for 2 min in the washing solution prior to visualization. HRP-3,3',5,5'Tetramethylbenzidine (TMB) (Molecular Probes) stain was used to visualize the reactive components. Membranes were digitized using a GS-800 Densitometer (Bio-Rad).

3.7. Western Blot on nitrocellulose (NC) membrane

After homogeneous SDS-PAGE, the proteins were electroblotted on NC membranes (Invitrogen) at 25 V using the tank transfer Xcell II Blot Module (Invitrogen), with Towbin 0.5X transfer buffer added with 5% methanol.

The proteins separated by the 4-12% gradient SDS-PAGE gels were electroblotted on NC membranes (Invitrogen) at 30 V, using the tank transfer Xcell II Blot Module (Invitrogen), with the NuPAGE Transfer buffer (Invitrogen) added with 5% methanol.

3.8. Preparation of the positive control for 4-hydroxy-2-nonenal (HNE) immunostaining

In order to evaluate the degree of single protein carbonylation due to binding to the oxidized lipids (indirect pathway of carbonylation), a specific immunostaining of the HNE adducts was performed on the NC membrane. A 4-hydroxy-2-nonenal dimethylacetal (HNE-DMA) hexane solution (Sigma) was used to obtain a positive HNE control. The HNE-DMA was evaporatedunder a gentle stream of nitrogen at room temperature to eliminate the hexane, and 400 μL of 1 mmol/L cold HCl at 4�C were added and stirred for 45 min. This solution was added to a mix of the three milk samples and used as the positive control.

3.9. HNE immunostaining

The NC membranes were incubated for 1 hour at room temperature in a blocking solution (PBS buffer containing 1% gelatin) (BioRad), and washed for 5 min in a washing solution containing PBS and 0.05% Tween-20. The membranes were then incubated overnight with the primary antibody solution, that consisted of a 1:5,000 dilution of the rabbit anti-HNE antiserum in an incubation solution containing PBS, 0.05% Tween-20 and 1% gelatine (Bio-Rad) at 4�C, while this passage was omitted in the negative control. The membranes were then washed twice for 5 min in the washing solution. They were incubated with a 1:2,500 dilution of the HRP-Goat Anti-Rabbit IgG Conjugate (Invitrogen) in the incubation solution for 2 hours at room temperature. The membranes were then washed twice for 5 min in the washing solution, and once for 5 min with PBS alone. The 4-chloro-1-Naphtol (Sigma) stain was used to visualize the reactive components. The membranes were digitized using a GS-800 Densitometer (Bio-Rad).

3.10. IgA Immunodetection

Four-mm slices were cut from the NC membranes of the gradient gels and incubated twice for 15 min at room temperature with a TBS blocking solution (50 mmol/L Tris, 150 mmol/L NaCl, pH 7.4) containing 0.3% Tween 20. The membranes were then incubated overnight at room temperature with the primary antibody IgA heavy chain rabbit anti-human Polyclonal Antibody Lifespan Biosciences diluted 1:3,000 with the incubation solution (TBS, containing 0.05% Tween-20 and 0.05% gelatin, Bio-Rad). The membranes were then washed 3 times for 5 min in a washing solution containing TBS and 0.05% Tween-20, incubated with a 1:5,000 dilution of alkaline phosphatase-labelled goat anti-rabbit IgG (PerkinElmer) in the incubation solution for 1 h at room temperature, and then washed 3 times for 5 min in the washing solution. The immunoblots were developed with the phosphatase substrate BCIP/NBT (5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium) (AP Conjugate substrate kit, Bio-Rad). The membranes were digitized using a GS-800 Densitometer (Bio-Rad).

3.11. Mass spectrometry

The bands of interest from SDS PAGE have been identified by means of mass spectrometry analysis. For MALDI TOF and LC-nano-ESI-ion trap analyses, the bands were cut and destained overnight with a solution of 50 mmol/L ammonium bicarbonate and 40% ethanol, then washed three times for 10 min with acetonitrile and dried in a Speedvac. After, the proteins bands were reduced with 50 �l of a solution of of DTT 200mM/100 mM NH4HCO3 for 1h at 56�C and then alkylated with 500 �l of a solution of 200mM iodoacetamide/100 mM NH4HCO3 for 45 min, RT. The proteins were then digested in gel with trypsin (Promega), as described by Hellmann and co-workers (38). For the MALDI-TOF mass spectrometry, 0.5 m L of each peptide mixture was applied to a target disk, and allowed to dry. Then 0.5 m L of HCCA matrix solution (α-cyano-4-hydroxycinnamic acid) was applied to the dried sample and allowed to dry again. Spectra of protein digests were obtained using a Bruker Ultraflex II MALDI-TOF/TOF mass spectrometer (Bruker Daltonics). Mascot (39) and MS-Fit (40) software packages were used to interpret the MS spectra through the peptide mass fingerprinting method (PMF) (41). Both software packages were searched against the NCBInr_20090123 database (42). The parameters used for the search were: S-carbamidomethyl derivate on cysteine as fixed modification, oxidation on methionine as variable modification and two missed cleavage sites for trypsin digestion. The peptide mass tolerance was 30 ppm.

Alternatively, the tryptic peptide mixtures were analyzed by means of LC-MS/MS using an Agilent-Technologies 1100 series nano HPLC, coupled to an Agilent XCT Plus ion trap fitted with a nano electrospray nebulizer. The chromatographic separations were run on a 150 x 0.075 mm C18 nanocolumn Zorbax 300SB (Agilent) using a linear 5-70% gradient of acetonitrile-0.1 % formic acid for 55 min, with a flow of 0.3 �L/min. The injection volume was 1 �L. The mass spectrometry parameters were: capillary voltage 1,600 V, fragmentation voltage 1.3 V. The Data Analysis software (Agilent) was used to elaborate the LC and MS/MS data. The Mascot search engine (39) was used to identifiy the proteins against the NCBInr_20090123 database (42). The parameters used for the search were: S-carbamidomethyl derivate on cysteine as fixed modification, oxidation on methionine as variable modification and two missed cleavage sites for trypsin digestion. The peptide mass tolerance was set to 0.5 Da and the fragment mass tolerance to 0.8 Da. The criteria used for protein identification were: 20 ion score with a protein Mascot score of more than 100 and at least two unique peptide sequences.

3.12. N-Terminal Sequencing

The bands of interest were excised and passively eluted as already described (43). The membranes were then microsequenced on a Procise 492 protein sequencer (Applied Biosystems). All the chemicals used in the procedure were from Applied Biosystems. The obtained amino acid sequences were compared with the protein sequences in the Uni-PROT database, using BLAST software (42).

3.13. Measurement of the lipase activity

The measurement of the lipase activity was performed by quantifying the reaction product p-nitrophenol (36). In a test tube, 50 μL of Raw, Holder, Flash and boiled (5 min at 95�C) unskimmed milk samples were mixed with 1 mL of 0.05 mol/L barbital buffer pH 7.6 (0.2 mol/L sodium barbital pH 7.6 with 0.2 mol/L HCl) and incubated at 37�C for 15 min. Then, 5 �L of substrate (50 mmol/L p-nitrophenyl butyrate in acetonitrile) were added, and the reaction mixtures were incubated at 37 �C for 10 min. Finally, 1 mL of Clarifying Reagent (Prolabo) was added to each tube, which was then stirred and incubated at 37 �C for 3 min. Absorbance at 420 nm was read within 15 min. The absorbance values were converted to μmol of p-nitrophenol using a standard curve in the 20 -200 μmole range. The boiled milk was used as a negative control to check the lipase inactivation, as reported by Williamson et al. (1978) (37) and Tully et al. (2001) (11). The determinations were replicated nine times. Statistical analysis was performed by a pairwise Student T-Test (p≤0.05).

3.14. Available lysine determination

The o-phthaldialdehyde (OPA) and N-α-acetyl-lysine were purchased from Sigma. All the reagents were of analytical grade, and Millipore Milli-Q water was used. The total available lysine content was measured using the OPA method (25,26). The OPA reagent (80 mg OPA dissolved in 1 mL ethanol 95%) was prepared immediately before use, and mixed with 99 mL of 0.05 mol/L sodium tetraborate buffer (pH 9.5) containing 5 mL of 20% SDS and 0.2 mL of 2-mercaptoethanol. A standard curve with a concentration ranging from 0.0125 to 0.375 mmol/L was obtained using N-α-acetyl-lysine as a reference compound, dissolved in 0.1 mol/L phosphate buffer pH 7.0 (27). The reference N-α-acetyl-lysine samples and 25 μL of unskimmed milk samples were added to 2 mL of the OPA reaction solution, briefly stirred, and left at room temperature for 2 min to incubate. The OD at 340 nm was measured against OPA reaction solution using an UV-VIS DU730 spectrophotometer (Beckman Coulter).

In order to eliminate the interference due to milk opalescence, blanks containing 25 μL of unskimmed milk in OPA reagent and not containing OPA (0.05 mol/L sodium tetraborate buffer pH 9.5, 1% SDS, 0.95% ethanol and 0.2% 2-mercaptoethanol) were used. The OD from each blank was subtracted from the OD obtained for the respective milk sample mixed with OPA-containing reaction buffer. This procedure allows the milk opacity due to both the casein micelles and the milk fat globules dispersion to be removed. The determinations were replicated seven times. The statistical analysis was performed by a pairwise Student T-Test and the results were considered statistically significant when associated to a probability of less than 5%.

4. RESULTS

A monodimensional SDS-PAGE of Raw, Holder and HTST milk samples was run, in order to discover the changes induced by pasteurization on the protein pattern. The resulting SDS-PAGEs and the corresponding grey-scale intensity chromatograms are shown in Figure 1. The HTST milk protein pattern appeared to be very similar to that of Raw milk, while the Holder pasteurized milk showed major differences. Heating milk at 62.5�C for 30 min reduced the intensities of the A, B and C bands. On the other hand, the E and F bands only appeared in the Holder milk and not in the other samples.

All the bands of interest were subjected to PMF by mass spectrometry analysis in order to identify the corresponding proteins. As shown in Table 1, the B and C bands were found to correspond to the oncofetal isoform of bile salt-stimulated lipase and to human lactoferrin, respectively. The A, E and F bands could not be identified.

The degree of protein carbonylation was assessed on the same homogeneous SDS-PAGE gels by DNP immunostaining. The results are shown in Figure 2. No aspecific reaction was detected in the control membrane (Figure 2, III). All the samples showed some degree of immunoreaction (Figure 2, II). Compared to the Raw and HTST milk, the Holder milk sample showed a reduced immunoreaction on the B and C bands, which correspond to the bands that showed a decreased intensity in the SDS PAGEs (Figure 1 and 2,I). In addition, the lipid oxidation of proteins was assessed by revealing the HNE adducts, through specific immunostaining. The obtained results are shown in Figure 3. The pasteurized milk samples suffered from a certain degree of lipid oxidation. The modified bands, named C and P, were identified as lactoferrin and β-casein (Table 2), respectively.

In order to characterize the protein profile more clearly at high and low molecular weights, all the milk samples were also subjected to a gradient SDS-PAGE. The resulting gels and chromatograms are shown in Figure 4. In the Raw milk, the C band was divided into two portions, one on the upper side (C₁) and one on the lower side (C₂). The C₁ and C₂ bands, and C band from the Holder milk, were all identified as human lactoferrin. Bands L, M, and N from Holder milk resulted to be fragments of β-casein (Table 3). The J and K bands, that completely disappeared in the Holder milk, were identified in the raw milk as components of the immune system (Ig λ-light chain variable region, MHC class I antigen) (Table 3). Finally, the O and Q bands, which were more intense in the Raw milk, detectable in the Flash milk and almost absent in the Holder milk, were identified through specific immunoreaction as human IgA (Table 3).

In order to assess whether the pasteurization method effected the level of lipase activity, a specific assay was conducted and the results are shown in Figure 5. The Holder milk showed a lack of lipase activity, while a comparable (p≥0.05) level of activity was detected for the Raw and HTST milk.

In addition, a specific assay was performed in order to investigate how the two pasteurization treatments effect the availability of lysine in milk. The mean contents of the available lysine of Raw, Holder and HTST milk samples are reported in Figure 6. The results indicate that the Holder milk had a significant (p≤0.05) higher content of available lysine.

5. DISCUSSION

Donor HM is the preferred feeding for preterm infants, when mother's own milk is not available in sufficient quantity. The supply of donor milk should come from established HM banks, which adhere to specific quality control guidelines (6-9). These guidelines (6-8) recommend that the Holder pasteurization method (62.5° , 30 min) is used to ensure the microbiological safety of donor HM. It has been shown that key nutritional and biological components, such as oligosaccharides, vitamins A, D and E, lactose, long chain PUFAs and the epidermal growth factor (which is important for intestinal maturation) are preserved by the Holder method (11, 13, 44-46). However, this process leads to the loss of some biologically active milk components: secretory IgAs, total IgAs, lactoferrin, lysozyme, lymphocytes, lipase, alkaline phosphatase, cytokines, and some growth factors. (10-14). The loss in these biological components in donor milk following pasteurization is a well-known problem, and often results in skepticism regarding its nutritional and immunological quality. Thus, the optimization of HM pasteurization has become one of the priorities of HM banks.

HTST pasteurization seems to be a valuable alternative to the Holder method as far as minimizing the loss of nutritional and immunological components while ensuring microbiological safety. Although there is no uniformity in the application of this method, with regards to the temperature (71�C to 75�C), duration (5-18.5 s), or the instrumentation, there is already a convincing body of evidence concerning its bacteriological and viral safety (19-21). Research is currently underway concerning its efficacy in retaining the nutritional and immunological quality of HM. The available data suggest that HTST pasteurization is better at preserving lactoferrin, IgA, IgG, and insulin like growth factors, and maintaining the antioxidant capacity of HM (13, 14, 19, 20).

5.1. Effect of pasteurization on lactoferrin

The present study is, to the authors' knowledge, the first to address the effects of different pasteurization methods, in terms of oxidation damage, on individual HM proteins. Human milk protein oxidation induced by thermal treatments, via the Maillard reaction, by direct oxidation on the amino acid residues, and by indirect reaction via lipid oxidation, has been investigated. The HM protein pattern of the HTST-treated milk and the raw milk resulted to be similar, whereas the Holder method substantially modified several proteins. The apparent decrease in lactoferrin quantity in Holder sample is likely to be due to the heat induced formation of carbonyls and the subsequent formation of disulfide-bonded high molecular aggregates which prevent lactoferrin from entering the SDS-PAGE gels and being visualized. This phenomena was previously reported for bovine milk (57,58), and could explain the apparent lower level of carbonylation on this protein in the Holder sample compared to both Raw and HTST milk. The decrease in lactoferrin in HM following pasteurization has been reported by several authors (11,49). Mata et al. (50) have reported a denaturation temperature of 67�C for lactoferrin as isolated in HM, with a peak at 71.85�C. Mata et al. have also confirmed the suitability of HTST pasteurization for partially retaining the iron-binding capacity of lactoferrin (although decreased by 25%), whilst a 20 min heat treatment at 80�C could reduce it by one half (50). Lactoferrin carbonylation has been observed in both raw and pasteurized HM. Lactosylation, by changing the protein conformation and making it less accessible to proteases, has been reported to partially prevent proteolysis on lactoferrin in bovine milk (51). This could also be expected to take place for HM lactoferrin, since about 10% of the ingested lactoferrin has been found intact in faeces of breast-fed infants (52).

5.2. Effect of pasteurization on bile salt-stimulated lipase

Bile salt-stimulated lipase is subjected to structural modifications which lead to a total loss of its enzymatic activity, as shown in Figure 1 and 5. The results demonstrate that the Holder method completely inhibits lipase, whereas the HTST method preserves both the integrity of the molecule (Figure 1) and its activity (Figure 5). In fact, as reported by Williamson and co-workers (37), even after only 1 min of pasteurization at 62.5�C, the lipase activity is completely lost. On the contrary, in the present experiment the HTST method based on a higher temperature (72�C) and shorter time (5-15 s) process, has been shown to preserve lipase activity and integrity.

5.3. Effect of pasteurization on components of the immune system

The HTST treated HM preserved the integrity of other native proteins, beside lactoferrin and bile salt-stimulated lipase, such as some components of the immune system (the J and K bands) (Figure 4, Table 3). The O and Q bands, which represent IgAs (Figure 4, Table 3), seem to be more sensitive to thermal treatment, being less intense in the heat-treated milk in comparison to the Raw HM. This is in agreement with what has been reported by Bjiorksten et al. (49) and Dhar et al. (20), who reported that a small increase in temperature causes a loss of IgA activity.

5.4. Effect of pasteurization on total available lysine content

As far as β-casein is concerned, carbonylation is not dramatically increased by either pasteurization method. The Holder method instead induces some protein degradation. Since caseins are considered to be the main source of available lysine in milk products (53), both these aspects - the formation of peptides and the low level of carbonylation - could explain the slightly higher availability of lysine residues in the Holder treated milk. The mean value of the total available lysine found in raw breast milk is consistent with a previous report (54). The slightly higher lysine content of the Holder treated milk found in our experiment is not consistent with a previous paper, which reported a strong reduction of lysine availability in thermal treated breast milk samples, and especially in those pasteurized by the Holder method (52).

In conclusion, different heat treatments have different impacts on HM proteins. The HTST treated HM resulted in a protein profile that was close to that of raw milk, with the preservation of the native forms of lactoferrin, and bile salt-dependent lipase, and to some extent of IgAs and some other components of the immunosystem. The HTST pasteurization method not only retained the integrity, but also the activity of the bile salt-stimulated lipase. This data could have important clinical implementations for donor milk use in terms of improved fat digestion and infection prevention for premature infants. The Holder treated HM seemed to have a higher content of bioavailable lysine. Briefly, HTST pasteurization seems to be better than the Holder method in retaining the protein profile, and some of the key biologically and immunologically active components of donor HM.

6. ACKNOWLEDGEMENTS

Cristina Baro and Marzia Giribaldi equally contributed to the paper. This research was supported by the Italian Association of Human Milk Banks, Milan, Italy.

7. REFERENCES

1. Y. Morales and RJ. Schanler: Human milk and clinical outcomes in VLBW infants: how compelling is the evidence of benefit? Semin Perinatol, 31, 83-88 (2007)
doi:10.1053/j.semperi.2007.02.002
PMid:17462492

2. G. Henderson, S. Craig, P. Brocklehurst and W. McGuire: Enteral feeding regimens and necrotising enterocolitis in preterm infants: a multicentre case-control study. Arch Dis Child Fetal Neonatal Ed, 94, F120-F123 (2009)
doi:10.1136/adc.2007.119560
PMid:17768154

3. J. Meinzen-Derr, B. Poindexter, L. Wrage, A.L. Morrow, B. Stoll and E.F. Donovan: Role of human milk in extremely low birth weight infants' risk of necrotizing enterocolitis or death. J Perinatol, 29, 57-62 (2009)
doi:10.1038/jp.2008.117
PMid:18716628    PMCid:2801431

4. L.M. Gartner, J. Morton, R.A. Lawrence, A.J. Naylor, D. O'Hare, R.J. Schanler and A.I. Eidelman: Breastfeeding and the use of human milk. Pediatrics, 115, 496-506 (2005)
doi:10.1542/peds.2004-2491

5. Academy of Breastfeeding Medicine Protocol Committee. ABM clinical protocol #3: hospital guidelines for the use of supplementary feedings in the healthy term breastfed neonate, revised 2009. Breastfeed Med, 4, 175-182 (2009)
doi:10.1089/bfm.2009.9991
PMid:19739952

6. Human Milk Banking Association of North America (HMBANA). Guidelines for the establishment and operation of a donated human milk bank. Sandwich, MA, HMBANA (2009)

7. Guidelines for the Establishment and Operation of Human Milk Banks in the UK. 3rd Edition. London: Royal College of Paediatrics and Child Health (2003)

8. G. De Nisi, A.M. Ambruzzi, S. Arslanoglu, E. Bertino, A. Biasini, G. Moro, C. Profeti, M.R. Spreghini and P. Tonetto. Linee guida per la costituzione e l'organizzazione di una banca del latte umano donato. New Magazine, Trento, Italy (2007)

9. B.T. Hartmann, W.W. Pang, A.D. Keil, P.E. Hartmann and K. Simmer: Australian Neonatal Clinical Care Unit. Best practice guidelines for the operation of a donated human milk bank in an Australian NICU. Early Hum Dev, 83, 667-673 (2007)
doi:10.1016/j.earlhumdev.2007.07.012
PMid:17913402

10. N.E. Wight: Donated human milk for preterm infants. J Perinatol, 21, 249-254 (2001)
doi:10.1038/sj.jp.7200533
PMid:11533843

11. DB Tully, F Jones and M.R.Tully: Donated milk: what's in it and what's not. J Hum Lact, 17, 152-155 (2001)
doi:10.1177/089033440101700212
PMid:11847831

12. A. Koenig, E.M. de Albuquerque Diniz, S.F. Barbosa and F.A. Vaz: Immunologic factors in human milk: the effects of gestational age and pasteurization. J Hum Lact, 21, 439-443 (2005)
doi:10.1177/0890334405280652
PMid:16280560

13. R. Goelz, E. Hihn, K. Hamprecht, K. Dietz, G. Jahn, C. Poets and M. Elmlinger: Effects of different CMV-heat-inactivation-methods on growth factors in human breast milk. Pediatr Res, 65, 458-461 (2009)
doi:10.1203/PDR.0b013e3181991f18
PMid:19127217

14. D. Silvestre, M. Miranda, M. Muriach, I. Almansa, E. Jareño and F.J. Romero: Antioxidant capacity of human milk: effect of thermal conditions for the pasteurization. Acta Paediatr, 97, 1070-1074 (2008)
doi:10.1111/j.1651-2227.2008.00870.x
PMid:18477059

15. R.J. Schanler: Mother's own milk, donated human milk, and preterm formulas in the feeding of extremely premature infants. J Pediatr Gastroenterol Nutr, 45, 175-177 (2007)
doi:10.1097/01.mpg.0000302967.83244.36
PMid:18185087

16. M.A. Quigley, G. Henderson, M.Y. Anthony and W. McGuire: Formula milk versus donated breast milk for feeding preterm or low birth weight infants. Cochrane Database Syst Rev, 17, CD002971 (2007)

17. C.A. Boyd, M.A. Quigley and P.Brocklehurst: Donated breast milk versus infant formula for preterm infants: systematic review and meta-analysis. Arch Dis Child Fetal Neonatal Ed, 92, F169-F175 (2007)
doi:10.1136/adc.2005.089490
PMid:16556615    PMCid:2675323

18. E. Bertino, F. Giuliani, L. Occhi, A. Coscia, P. Tonetto, F. Marchino and C. Fabris: Benefits of donated human milk for preterm infants: current evidence. Early Hum Develop, 85, 9-10 (2010)
doi:10.1016/j.earlhumdev.2009.08.010
PMid:19800748

19. R.M. Goldblum, C.W. Dill, T.B. Albrecht, E.S. Alford, C. Garza and A.S. Goldman: Rapid high-temperature treatment of human milk. J Pediatr, 104, 380-385 (1984)
doi:10.1016/S0022-3476(84)81099-9

20. J. Dhar, J. Fichtali and B.J. Skura: Pasteurization efficiency of a HTST system for human milk. J Food Sci, 61, 569-573 (1996)
doi:10.1111/j.1365-2621.1996.tb13160.x

21. K. Hamprecht, J. Maschmann, D. Müller, K. Dietz, I. Besenthal, R. Goelz, J.M. Middeldorp, C.P. Speer and G. Jahn: Cytomegalovirus (CMV) inactivation in breast milk: reassessment of pasteurization and freeze-thawing. Pediatr Res, 56, 529-535 (2004)
doi:10.1203/01.PDR.0000139483.35087.BE
PMid:15295084

22. B. Lönnerdal: Bioactive proteins in human milk: mechanisms of action. J Pediatr, 156, 26-30 (2010)
doi:10.1016/j.jpeds.2009.11.017
PMid:20105661

23. M.A.J.S. Van Boekel: Effect of heating on Maillard reactions in milk. Food Chem, 62, 403-414 (1998)
doi:10.1016/S0308-8146(98)00075-2

24. F. Fenaille, V. Parisod, J.C. Tabet and P.A. Guy: Carbonylation of milk powder proteins as a on sequence of processing conditions. Proteomics, 5, 3097-3104 begin_of_the_skype_highlighting              3097-3104      end_of_the_skype_highlighting (2005)
doi:10.1002/pmic.200401139
PMid:16038017

25. C.C. Goodno, H.E. Swaisgood and G.I. Catignani: A fluorimetric assay for available lysine in proteins. Anal Biochem, 115, 203-211 (1981)
doi:10.1016/0003-2697(81)90547-9

26. F.C. Church, H.E. Swaisgood, D.H. Porter and G.I. Catignani: Spectrophotometric assay using o-phthaldialdehyde for determination of proteolysis in milk and isolated milk proteins. J Dairy Sci, 66, 1219-1227 (1983)
doi:10.3168/jds.S0022-0302(83)81926-2

27. C. Dinnella, M.T. Gargaro, R. Rossano and E. Monteleone: Spectrophotometric assay using o-phtaldialdehyde for the determination of transglutaminase activity on casein. Food Chem, 78, 363-368 (2002)
doi:10.1016/S0308-8146(02)00109-7

28. U.K. Laemmli: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685 (1970)
doi:10.1038/227680a0
PMid:5432063

29. G. Candiano, M. Bruschi, L. Musante, L. Santucci, G.M. Ghiggeri, B. Carnemolla, P. Orecchia, L. Zardi and P.G. Righetti. Blue silver: a very sensitive colloidal Coomassie G-250 staining for proteome analysis. Electrophoresis, 25, 1327-1333 (2004)
doi:10.1002/elps.200305844
PMid:15174055

30. U. Panzenboeck, S. Raitmayer, H. Reicher, H. Linder, O. Glatter, E. Male and W. Sattler: Effects of reagent and enzymatically generated hypoclorite on physicochemical and metabolic properties of high density lipoproteins. J Biol Chem, 272, 29711-29720 (1997)
doi:10.1074/jbc.272.47.29711
PMid:9368040

31. H. Buss, T.P. Chan, K.B. Sluis, N.M. Domigan and C.C. Winterbourn: Protein carbonyl measurement by a sensitive ELISA method. Free Radic Biol Med, 23, 361-366 (1997).
doi:10.1016/S0891-5849(97)00104-4

32. H. Towbin, T. Staehelin, and J. Gordon: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA, 76, 4350-4354 (1979)
doi:10.1073/pnas.76.9.4350

33. C.E. Robinson, A. Keshavarzian, D.S. Pasco, T.O. Frommel, D.H. Winship and E.W. Holmes: Determination of protein carbonyl groups by immunoblotting. Anal Biochem, 266, 48-57 (1999)
doi:10.1006/abio.1998.2932
PMid:9887212

34. C.C. Conrad, J.M. Talent, C.A. Malakowsky and R.W. Gracy: Post-electrophoretic identification of oxidized proteins. Biol Proced Online, 239-245 (2000)

35. M.A. Mansfield. Rapid immunodetection on Polyvinylidene Fluoride membrane blots without blocking. Anal Biochem, 229, 140-143 (1995)
doi:10.1006/abio.1995.1391
PMid:8533884

36. G. Humbert, M.F. Guingamp and G. Linden: Method for the measurement of lipase activity in milk. J Dairy Res, 64, 465-469 (1997).
doi:10.1017/S0022029997002288
PMid:9275261

37. Hernell O, Olivecrona T. Human milk lipases II. Bile salt-stimulated lipase. Biochim Biophys Acta, 369, 234-244 (1974).
PMid:4429695

38. U. Hellmann, C. Wernstedt, J. Gonez and C.H. Heldin: Improvement of an in-gel digestion procedure for the micropreparation of internal protein-fragments for amino-acid sequencing. Anal Biochem, 224, 451-455 (1995)
doi:10.1006/abio.1995.1070
PMid:7710111

39. http://www.matrixscience.com/

40. http://prospector.ucsf.edu/

41. D.J. Pappin, P. Hojrup and A.J. Bleasby: Rapid identification of proteins by peptide-mass fingerprinting. Curr Biol, 3, 327-332 (1993)
doi:10.1016/0960-9822(93)90195-T

42. http://blast.ncbi.nlm.nih.gov/Blast.cgi

43. A. Reuter, D. Fortunato, L. Perono Garoffo, L. Napolitano, S. Scheurer, M.G. Giuffrida, S. Vieths and A. Conti: Novel isoforms of Pru av 1 with diverging immunoglobulin E binding properties identified by a synergistic combination of molecular biology and proteomics. Proteomics, 5, 282-289 (2005)
doi:10.1002/pmic.200400874
PMid:15593144

44. E. Bertino, G.V. Coppa, F. Giuliani, A. Coscia, O. Gabrielli, G. Sabatino, M. Sgarrella, T. Testa, L. Zampini, C. Fabris: Effects of Holder pasteurization on human milk oligosaccharides. Int J Immunopathol Pharmacol, 21, 381-385 (2008)
PMid:18547467

45. D. Van Zoeren-Grobben, J. Schrijver, H. Van Den Berg and H.M. Berger: Human milk vitamin content after pasteurisation, storage, or tube feeding. Arch Dis Child, 62, 161-165 (1987)
doi:10.1136/adc.62.2.161
PMid:3103547    PMCid:1778245

46. T.R. Henderson, T.N. Fay and M. Hamosh: Effect of pasteurization on long chain polyunsaturated fatty acid levels and enzyme activities of human milk. J Pediatr, 132, 876-878 (1998)
doi:10.1016/S0022-3476(98)70323-3

47. H.A. Patel, H. Singh, S.G. Anema and L.K. Creamer: Effects of heat and high hydrostatic pressure treatments on disulfide bonding interchanges among the proteins in skim milk. J Agric Food Chem, 54, 3409-3420 (2006)
doi:10.1021/jf052834c
PMid:16637702

48. G. Brisson, M. Britten and Y. Pouliot: Heat-induced aggregation of bovine lactoferrin at neutral pH: effect of iron saturation. Int Dairy J, 17, 617-624 (2007)
doi:10.1016/j.idairyj.2006.09.002

49. B. Björkstén, L.G. Burman, P. De Château, B. Fredrikzon, L. Gothefors and O. Hernell: Collecting and banking human milk: to heat or not to heat? Br Med J, 281, 765-769 (1980)

50. L. Mata, L. Sánchez, D.R. Headon and M. Calvo: Thermal denaturation of human lactoferrin and its effect on the ability to bind iron. J Agric Food Chem, 46, 3964-3970 begin_of_the_skype_highlighting              3964-3970      end_of_the_skype_highlighting (1998)
doi:10.1021/jf980266d

51. T.K. Dalsgaard, J.H. Nielsen and L.B. Larsen: Proteolysis of milk proteins lactosylated in model systems. Mol Nutr Food Res, 51, 404-414 (2007)
doi:10.1002/mnfr.200600112

52. B. Lönnerdal: Nutritional and physiologic significance of human milk proteins: Am J Clin Nutr, 77, 1537S-1543S (2003)
PMid:12147341

53. A. Scaloni, V. Perillo, P. Franco, E. Fedele, R. Froio, L. Ferrara and P. Bergamo: Characterization of heat-induced lactosylation products in caseins by immunoenzymatic and mass spectrometric methodologies. Biochim Biophys Acta, 1598, 30-39 (2002)
PMid:16614484

54. D. Silvestre, E. Ferrer, J. Gayá, E. Jareño, M. Miranda, M. Muriach and F.J. Romero: Available lysine content in human milk: stability during manipulation prior to ingestion. Biofactors, 26, 71-79 (2006)
doi:10.1002/biof.5520260107
PMid:7427442    PMCid:1714023

Key words: Donor Milk, Very Low Weight Infants, Human Milk Banks, Proteins Profile, Pasteurization, Lipase, Lactoferrin, Mass Spectrometry