Heme biosynthesis in human breast cancer--mimetic "in vitro" studies and some heme enzymic activity levels.

1. Porphyrin biosynthesis from delta-aminolevulinic acid (ALA) was investigated using the technique of tissue explant cultures, in both human breast cancer and its original normal tissue. 2. The activity of ALA-dehydratase, porphobilinogenase and uroporphyrinogen decarboxylase was directly determined in both tumor and normal mammary tissues. 3. Porphyrin synthesis capacity of human breast carcinoma was 20-fold enhanced, as compared with normal tissue, at least between the stages of porphobilinogen and coproporphyrinogen formation. 4. The activity of the three enzymes examined was always lower in normal tissue than in tumoral tissue. 5. Present findings show that porphyrin biosynthesis is increased in breast cancer tissue.


INTRODUCTION
Human breast cancer is one of the most frequent type of malignant tumors found in females, leading to the highest rate of mortality as compared with that due to any other tumors affecting this sex. In practice the treatment of choice is surgery followed by radiotherapy, chemotherapy and hormone therapy. More recently, other treatment reported to be rather effective, has been photodynamic therapy (PDT) (Dougherty et al., 1979), consisting in the administration of the so-called hemotoporphyrin derivatives (HpD) or Photofrin II a more purified product already approved by the FDA for its use in humans. PDT is based on the property of certain porphyrins to efficiently accumulate in neoplastic tissue (Rassmussen-Taxdall et al., 1955;Lipson et al., 1961;Gregorie et al., 1968;Dougherty et al., 1978;Gomer and Dougherty, 1979), and the photodynamic properties of these compounds, which would then allow selective killing of the tumorous cells with little damage to the surrounding normal cells.
A good deal of research on this interesting area has been mostly concerned with the tissular distribution of porphyrins after injection (Gomer and Dougherty, 1979;Bugelski et al., 1981) and the subcellular localization of accumulated porphyrins (Kessel, 1981;Sandberg and Romslo, 1981;Jori et al., 1986). However, the reasons for this porphyrins discrimination between normal and tumor cells are not yet known.
As a first approach to the problem, we consider of value to investigate comparatively the functionality *Dedicated to the memory of Dr Torben K. With. tTo whom all correspondence should be addressed at: Viamonte 188 1, 10" "A", 1056 Buenos Aires, Republica Argentina.
of the heme pathway in both neoplastic and the corresponding normal tissue, from human.
From preliminary studies carried out to establish the optimum experimental conditions, it was found that porphyrin synthesis from the precursor 6 -aminolevulinic acid (ALA) is significantly enhanced in explant cultures of human breast carcinoma (Navone et al., 1988).
On the other hand, it is worth recalling that alterations in the drugs detoxifying system, particularly with reference to some heme proteins of the so-named phase I, have been described (Denk et al., 1980;Farber, 1984;Stout and Becker, 1986).
These findings allow us to speculate about the existence of deviations from normal in the heme pathway operating in tumor cells.
So far, information available in this field is only partial and most of it comes from studies using experimental models Becker, 1986, 1987;Bonkowsky et al., 1973).
We will present here results on our investigation on the biosynthesis of porphyrins by human breast cancer tissues in comparison with the original normal tissue, using the novel technique of tissue explant cultures and on the other hand we will report on the direct measurement of the levels of some of the heme enzymes in the same kind of tissues. From these findings we can ensure that in human breast carcinoma, porphyrin biosynthesis capacity is notoriously increased, at least from the stage of porphobolinogen (PBG) formation up to that of coproporphyrinogen, when compared with its original normal tissue.

Tissues
Tissues came from the Instituto National de Oncologia "Angel H. Roffo". Human breast carcinoma was used. As a matched control, normal mammary tissue from the same patient was used. Samples were obtained after surgical therapy from female adult patients (2666 years old), with the diagnosis of carcinoma and receiving no treatment; they were kept at 0°C and processed within 4 hr after having been obtained.

Histology
Rutinary histological examination was carried out after fixing in 10% formol, paraffin inclusion, cut and then tinction with hematoxiline and eosine. Every tumor was classified according to its histological malignancy following the recommended international classification for tumors (WHO).

Explant tissue cultures
Fractions of 50-60 mg were kept in the designed culture chamber as described by Vazquez et al. (1987) and run under the conditions previously established (Navone et al., 1988).

Porphyrins determination
Porphyrins present in the medium were determined fluorometrically by the procedure and using arbitrary units of fluorescence (FRU) as described and defined by . Because of possible cellular lysis, results were also normalized taking into account the degree of lysis, and defined a lysis units (LU) as a function of lactate dehydrogenase (LDH) activity in the medium as already reported (Vbzquez et al., 1986). Porphyrin accumulation was therefore measured only in the culture medium and calculated as FRU/mg tissue x LU (Buzaleh et al., 1988).

Homogenates
Enzyme activities were measured in homogenates pre- It should be noted that in the case of porphyrins formed, they are expressed as FRU, so SA will be FRU/mg protein.

Studies on tissue explant cultures
Serial and comparative studies were performed in samples from 7 patients. It was found that in an average porphyrin formation was enhanced ca 20 fold in the neoplastic tissue when compared with the original normal tissue (Table 1). Before and after incubation, histology was examined in each tissue; it was observed that the nuclear area was well preserved instead the cytoplasm had lost definition probably due to cellular membrane damage, it must be noted that, as indicated in Materials and Methods, to account and correct for the probable cellular lysis, LDH activity was measured in the maintenance medium (Vazquez et al., 1986). The number of parenquimatous cells in both neoplastic and normal tissue was also quantified by means of a histological evaluation. Thus, it resulted that the density of cellular population in the former tissue in some cases corresponded to 80% of the sample, while in those samples showing prevalence of fibrous reached to only 20%. In normal mammary tissue values were between 10 and 30%. It should also be noted that no correlation whatsoever was found between the grade of histological malignancy and the magnitude of porphyrin biosynthesis enhancement.

Enzymic activity levels
The above findings, although undoubtedly indicating increased capacity for heme synthesis by the neoplastic tissue would only be an indirect measure of the phenomenon; direct quantification of enzyme activity would reflect more definitely and possibly individualize the changes occurred if any. So, ALA-D, PBGase and URO-D were determined in both neoplastic and normal tissue.

ALA-D
It was measured in samples from 5 patients (Table 2). In every pair activity was higher in tumor than in normal cells, activity of the latter was between 6 and 60% of the corresponding matched malignant cells. However, the variability among different samples was so great that signification tests were negatives. Moreover, ALA-D is usually in excess over the other enzymes of the pathway, therefore we will not risk any definite comment about its relative increased activity in tumor cells, as yet.  leukemia and lymphoma (Lahav et al., 1987).
*Expressed as FRU/mg protein (see Materials and Methods).
All abbreviations as in Table I.
Acknowledgements-This work was supported by grants from the National Research Council (CONICET). University of Buenos Aires and Banco de la Nation Argentina. Samples from 4 patients were examined (Table 3).  T, Tumor; L, liver; B, brain; K, kidney. IJ, Normal mice; q , tumor bearing mice. Each value represents the mean+SD obtained from 5 different mice of each type. Experimental details as well as enzyme units are described in the text.
In analogy with PBGase, tumoral HMB-S also showed values similar to those of liver and kidney [Fig. Al(c)]. As to the ratio between HMB-S and PBGase it was nearly 2 and approximately the same in all the studied tissues.
Tumoral URO-D activity as already discussed for PBGase and HMB-S was rather the same of that of liver and kidney [ Fig. Al(d)].
To the best of our knowledge studies on these enzymes in tumors are scarce and were carried out only in human specimens. Significant diminished ALA-D activity was found in a variety of tumors from human gut, when compared with normal tissues (Rasetti et al., 1967). However these same authors observed increased porphobilinogen biosynthesis in tumor tissues from lung and uterus as compared with the corresponding normal specimens (Rasetti et al., 1967).
In leukocytes from patients suffering either acute or chronic myelocytic leukemia 3%5-fold decreased ALA-D activity was detected in contrast with normal and mature lymphocytes and granulocytes (Takaku and Wada, 1968). HMB-S activity from peripheral lymphocytes was higher in patients with malignant lymphoproliferative diseases such as chronic lymphocytic leukemia and lymphoma than in normal individuals (Lahav et al., 1987).
Taking into account the above results coming from our and other groups, we conclude that a partially altered enzyme pattern could exist in neoplastic tissue as compared with a high heme producing tissue such as liver. Changes observed for ALA-D activity were quite different depending on the original tissue. It is interesting to note that HMB-S activity seems to be enhanced in all neoplastic tissues examined so far. Furthermore, the results reported here are also in agreement with studies carried out with human breast carcinoma.