Molecular Identification of Falciparum Malaria and Human Tuberculosis Co-Infections in Mummies from the Fayum Depression (Lower Egypt)

Due to the presence of the lake Quarun and to the particular nature of its irrigation system, it has been speculated that the Fayum, a large depression 80 kilometers south- west of modern Cairo, was exposed to the hazards of malaria in historic times. Similarly, it has been speculated that, in the same area, also human tuberculosis might have been far more widespread in the antiquity than in its recent past. If these hypotheses were confirmed, it would imply that frequent cases of co-infection between the two pathogens might have occurred in ancient populations. To substantiate those speculations, molecular analyses were carried out on sixteen mummified heads recovered from the necropolis of Abusir el Meleq (Fayum) dating from the 3rd Intermediate Period (1064- 656 BC) to the Roman Period (30 BC- 300 AD). Soft tissue biopsies were used for DNA extractions and PCR amplifications using well-suited protocols. A partial 196-bp fragment of Plasmodium falciparum apical membrane antigen 1 gene and a 123-bp fragment of the Mycobacterium tuberculosis complex insertion sequence IS6110 were amplified and sequenced in six and five of the sixteen specimens, respectively. A 100% concordance rates between our sequences and those of P. falciparum and M. tuberculosis complex ones were obtained. Lastly, concomitant PCR amplification of P. falciparum and M. tuberculosis complex DNA specific fragments was obtained in four mummies, three of which are 14 C dated to the Late and Graeco-Roman Periods. Our data confirm that the hydrography of Fayum was extremely conducive to the spread of malaria. They also support the notion that the agricultural boom and dense crowding occurred in this region, especially under the Ptolemies, highly increased the probability for the manifestation and spread of tuberculosis. Here we extend back-wards to ca. 800 BC new evidence for malaria tropica and human tuberculosis co-occurrence in ancient Lower Egypt.


Introduction
Tuberculosis (TB) and malaria, two of the most ancient and deadly diseases of mankind, have ravaged human communities since the beginning of civilization and remain a major global health problem in the 21 st century [1,2]. TB causes ill-health among millions of people each year and ranks as the second leading cause of death of adults from an infectious disease worldwide, after the human immuno-deficiency virus (HIV). The latest World Health Organization report indicate that there were almost 9 million new cases in 2011 and 1.4 million TB deaths [1].
Malaria is the 5 th cause of death from infectious diseases worldwide after respiratory infections, HIV/AIDS, diarrheal diseases and tuberculosis. In 2011, there were 1.2 million malaria deaths globally and malaria is recognized as the 2 nd leading cause of death from infectious diseases in Africa, after HIV/AIDS [2][3]. In many parts of sub-Saharian Africa, the geographic overlap between TB and malaria is extensive and co-infection with TB and malaria is likely to be common [4][5].
Several investigations aimed at tracing the origins and frequencies of malaria and TB back have been carried out on Egyptian mummified remains over the past 25 years [6][7][8][9][10][11][12][13][14][15]. Tuberculosis DNA has long been recognized in mummies from Upper Egypt dating to different historical periods [7][8][9][10][11][12][13]. Mycobacterium tuberculosis (MTB) complex DNA was successfully amplified and sequenced in populations from Thebes-West (Upper Egypt) dating to the New Kingdom (c.a. 1550-1000 BC) [7,[9][10][11][12]. Where additional characterization was possible, it was shown that human lineages of MTB complex were present: M. tuberculosis and possibly M. africanum but not M. bovis [12]. Molecular signatures of TB infection were found in individuals with and without macromorphological typical signs of tubercular spondylitis and belonging to different age classes [9]. It was speculated that, in the affected populations, there was a relatively low life expectancy and that this may have resulted from a considerable proportion of chronic infections by various pathogenic organisms such as tuberculosis and other parasitoses (i.e. malaria, leishmaniasis, schistosomiasis and other worm infections) [9,12].
Similarly, but to a lesser extent, also Plasmodium falciparum ancient DNA was identified in mummified skeletons from Thebes-West dating from the New Kingdom to the Late Period (1500-500 BC) [14] and in 18 th Dynasty royal mummies [15]. However, no cases of malaria and tuberculosis co-infections were reported. Furthermore, most of the studies carried out till the present day have focused on the impact of infectious diseases on ancient populations from Upper Egypt whereas less attention has been paid to populations from Lower Egypt. Stimulated by the notion that malaria and TB were rampant in 19 th and early 20 th century Fayum [16], the present studys aim was to detect and characterize P. falciparum and MTB complex DNAs in sixteen mummified heads from Lower Egypt and, eventually, to verify, the existence of co-infections.
Here we not only extend back-wards to ca. 800 BC evidence for single malarial and mycobacterial infections but provide new evidence of falciparum malaria/tuberculosis co-occurrence in individuals from ancient Fayum ( Table 1).
In addition, the parasite surface protein antigen MSP1 allele was successfully genotyped in three of six malaria-positive mummies (50%) (Figure 1b-here 1564-& Table 1). More specifically, DNA extracts from the positive AMA1 samples 1554, 1564 and 1622 showed amplification with K1 allele-specific primers whereas RO33 and MAD20 failed to amplify any of the 16 mummies tested ( Table 1). This K1 allele from mummy 1564 was exemplary cloned into E. coli Top10 competent cells. Direct sequencing of the amplified clones using M13 primers showed eight unique K1 alleles with varying length in the repeat sequence ( Figure 1c).

Discussion
Since 1998 there have been no officially reported autochtonous cases of malaria in the Fayum Governatorate or elsewhere in Egypt [17]. Before the eradication of malaria from Egypt, high levels of the infection appeared to have been limited to certain parts of the country and to be strictly linked to its geology [16]. Due to the presence of the Birket Quarun, also known as Lake Moeris (which is fed by Bahr Yusuf), and to the particular nature of its irrigation system (i.e. a system of a network of smaller canals), the Fayum was highly exposed to the hazards of malaria in its recent past.
In the third quarter of the 19 th century and early 20 th century, especially after the third land reclamation carried out under Muhammad Ali's reign (1805-1848), this area was reputed as the most malarial region outside the northern Delta and several cases of malarial and tubercular co-infections were registered in populations from this area [16].
Since the size of the lake and its water-level changed drastically over the past few millennia, it has been speculated that Fayum was even more conducive to malaria in historic times than in its recent past [16]. Following Sheidel [16], ''the manifestations of malaria that used to be endemic in the Mediterranean -quotidian, tertian and quartan fevers, associated with Plasmodium falciparum, vivax, and malariae -are explicitly attested in charms from Graeco-Roman Egypt. Hints at the occurrence of malaria can already be found in earlier periods, such as a warning in a temple inscription at Denderah not to leave home after sunset in the weeks following the inundation (when mosquitoes would have proliferated), reported strategies of mosquito evasion that are typical of malarial areas and, possibly, reference to the ''disease of the three days''.
The same type of malaria was still common in Egypt in the 1930s, suggesting continuity in the long term. Comparative evidence makes it seem likely that the marshy Fayum would have been particularly exposed to endemic malaria''.
Paleoenvironmental and paleoclimatic studies give us clues into the reconstruction of ancient Fayum's landscapes [18][19][20][21]. Fluctuations in Neolithic times led to the success or demise of various agriculture or fishing communities on the lake shore [16,[18][19][20][21]. Before the Dynastic times the lake level was high (18-20 meters above the sea level (masl)) and covered most of the extensive depression. A general recession set in around the beginning of the Dynastic times (ca. 2880 BC). Around 2200 BC extensive dry conditions occurred (lower rainfalls) which determined Fayum's lake receding levels. During the Old Kingdom (2663-2195 BC), the lake might have been as low as 2 meters below the sea level and no longer in free communication with the Nile [16][17][18][19][20][21][22].
In response to falling flood levels, the first land reclamation of Fayum occurred in the Middle Kingdom under the reigns of the 12 th Dynasty ''engineering kings'' [Senwosret II (1900-1880 BC) and Amenemhat III (ca 1880-1808 BC)] [16,22,23]. Although there's no archaeological or textual evidence that sheds light on the exact extent of the works carried out in this period, this hydraulic project included the dredging of a branch of the Nile through the opening and widening of the Hawara canal, now called Bahr Yusuf, to allow waters to re-flow into the Fayum depression [16][17][18]22].
Water flow was regulated to maintain the lake level at around 14 to 16 meters above the sea level in order to reclaim a vast area for agriculture. The locks that regulated the amount of water flowing into the depression are still located at the right side of the el-Lahun corridor [22]. A connected primitive canal system was created in order to irrigate the fields [16,22]. At a time after 1800 BC, due to unexpected increase in flow discharge, the level of the lake raised considerably (ca 18 to 20 masl) and most of the areas occupied by the Middle Kingdom installations were drawn [16,20]. The dam was breached and the Fayum depression became once more uninhabitable. The lake still stood at around the same level or even higher when Herodotus saw it in the mid-5 th century BC and apparently remained that way until early Ptolemaic times [16,22].
The second large scale reclamation of the Fayum started under the Ptolemies [Ptolemy I Soter (310-282 BC) and Ptolemy II Philadelphos (282-246 BC)] [22]. It probably took as long as 30 years to lower the level of the lake to two meters below sea level reaching the same level it had during the Old Kingdom.
The Lahun embankment (a five kilometers embankment from the northern side of the Lahun Gap at al-Lahun) constructed under the Ptolemies was used to divert the annual influx of Nile water. Lake Quarun, therefore, dropped to 5 meters below the sea level and the newly exposed land was colonized by Macedonian soldiers [16,22]. The campaign of land reclamation appears to have started in the eastern area of the Fayum, in the meris of Herakleides, then continued in the southern area, corresponding to the meris of Polemon and, lastly, into the western area in the meris of Themistos [22].
This project added about 1.200 square kilometers of fertile land to the Fayum and attracted settlers, many of them from abroad. The development project led to an agricultural boom in the region with large settlement programs and the founding of many new towns. Hundreds of settlements were created [16].
During the Roman period, the Fayum became the granary of Egypt and Rome. The lake continued to shrink during the Roman Period (35 BC-AD 385) and declined from -7 meters in the 2 nd century AD to -17 or lower in the 3 rd century BC [22]. The whole canal system continued to be used, the agricultural productivity especially of wheat and barley was further optimised. Settlements formed with periods of high population density. The most reasonable guess for the overall population density in the Roman Period is about 200-300 people per km squared [16].
The reclamation of wetlands for cultivation and the creation of a system of canals for the irrigation of crops were apparently the key events which led to the exacerbation of the risk of malaria in this area.
Modern studies focusing on the linkage of land reclamation for cultivation and malaria's spread in east Africa show that the cultivation of swampy depressions or swamp valley bottoms, as result of a rapid population and demand for more food, changes the local ecology [24,25]. The wetlands are originally covered with natural papyrus, which limits the breeding of Anopheles (the vector) because of the dense vegetation and the oil layer. The elimination of papyrus and the reclamation of the swamps leads to an increase in temperatures promoting breeding of mosquitoes and, therefore, by increasing malaria transmission. If the irrigation system lacks efficient drainage, water accumulation is facilitated and, in turns this provides a breeding ground for mosquitoes. On the average, all malaria indexes (i.e. mosquitoes densities, biting rates, sporozoite rates) show to be much higher near cultivated swamps than near natural swamps [24,25]. If the epidemiological relationship between increased malaria transmission and intensified crops cultivation is explored, the results show that the incidence of malaria is about was ten times higher in cereals-cultivation areas than in areas with less cereals (wheat pollens provide nutrition for larval anopheline mosquitoes). This implies that the intensity of crops cultivation is associated with exacerbated human risk of malaria. Also irrigation exposes non-immune populations in areas of unstable malaria transmission to high risk of acquiring the disease [24,25].
For ancient Fayum's populations, we hypothesize that reclamation of lands for cultivation and the constant contact between malaria's vectors and human beings might have determined an increased risk of acquiring malaria. Furthermore, dense crowding of people may have been a significant factor also for the manifestation and spread of human tuberculosis and the risk of contracting concomitant infections might have been, therefore, augmented.
We confirm, using modern methods, that malaria and tuberculosis were endemic to the Fayum Depression from ca. 800 BC until the Roman Period ( Table 1). High frequencies of malaria and tuberculosis co-infections (4/16) (25%) were identified in pre-adolescents (2 cases) and young adults (2 cases) which appeared to be, as it occurs in modern populations, the most affected age-classes.
Modern epidemiology shows that, in areas of the world where tuberculosis and malaria are co-endemic, malaria infection affects severely ill TB patients who are already compromised by malnutrition, deprived immunity and disseminated disease. Interactions between TB and malaria have been demonstrated in vitro and in vivo: Plasmodium falciparum modulates Mycobacterium tuberculosis infection [26] and malaria has been shown to exacerbate mycobacterial infection [27].
The reasons for this are not completely explored but seem to involve parasite-parasite interaction and host-parasite interaction: malaria causes a further depression in immunity through a qualitative and quantitative defect in T lymphocytes, mainly CD8+ that are necessary for anti-mycobacterial response, and through a degeneration of the cytokine cascade [28,29]. Moreover, the respiratory distress that is frequent during acute malaria both in children (due to metabolic acidosis) and adults (due to pulmonary edema and to acute respiratory distress syndrome) can worsen the respiratory efforts related to tuberculosis [30,31]. Therefore, given the multiple interaction between malaria and TB, there's a mutual effect in increasing mortality [32]. For the individuals we analyzed, we can hypothesize that P. falciparum infection played a significant role in increasing the incidence of reactivation of latent tuberculosis in adults or primary active tuberculosis in children and pre-adolescents, as it occurs in modern day populations.
On the basis of the results obtained so far, we feel confident to develop an 'ancient pathogens genome project' encouraged wholly by the improvement in sequencing technology well suited to ancient DNA research. This will allow us chronologically review pathogens evolution through their genomic sequences, compared with modern isolates and verify the evolutionary hypothesis about their origins. Many ongoing ancient DNA research aided by the ever improving next generation sequencing techniques in terms of throughput and analysis tools will provide a link to better insight into assessments of diseases in the past, their emergence and transmission.

Materials and Methods
We analyzed soft tissue biopsies taken from 16 mummified heads belonging to the collection of the Institute of Pre-and Protohistory, Department of Early Prehistory and Quaternary Ecology Division of Paleoanthropology (Tübingen, Germany) ( Table 1). Muscle biopsies (1 cm squared) from each head were taken from the inner neck region.
The heads (the rest of the bodies is missing) were recovered from the necropolis of Abusir el-Meleq in the Fayum Depression (Lower Egypt) at the beginning of the 20 th century [31,32] and date from the 3 rd Intermediate Period (1064-656 BC) to the Roman Period (30 BC-300 AD) [23,33].

Authentication
Two ancient DNA laboratories dedicated for DNA extraction, PCR and Post-PCR analyses were established at the Department of Medical Genetics and Applied Genomics, University of Tübingen. No analyses with malaria or tuberculosis clinical specimens have ever been carried out in any of these laboratories. In addition, isolated replication analyses have been performed in another ancient DNA laboratory in Munich. DNA extraction, PCR and post-PCR analyses were performed separately in a dedicated room for each work. All reagents purchased were certified DNAase and RNAase-free molecular grade chemicals or were autoclaved if applicable. Disposable gloves, coats, and mouth masks were worn during all different procedures and changed frequently. Sterile aerosol protection filter tips (Biozym, Oldendorf Germany) were used in all experiments. Extraction blanks as well as negative PCR controls without template were included for every DNA extractions and PCR amplification. Replicating experiments were performed employing different laboratory space/equipment and enzyme/water aliquots and data were further verified by PCR-cloning and sequencing.

DNA Extraction
Sampling from sixteen mummies ( Table 1), DNA extraction and purification were performed following the protocol described earlier with modifications [34]. For DNA extraction, approximately 50-100 mg of tissue powder was placed in a 2 mL tube, 800 ml of DNA lysis buffer consisting of 10 mM Tris-HCL, pH 8,0, 10 mM NaCl, 2% SDS and 200 mg/mL proteinase K enzyme and 800 ml of phenol were added and mixed well. The solution was incubated overnight at 37uC incubator on a shaker. DNA was purified by chloroform and precipitated by addition of 1/10 volume of 3 M sodium acetate (pH 5.2) and an equal volume of chilled isopropanol. The pellets were washed with 70% ethanol and subsequently air-dried. All pellets were dissolved in nucleasefree water (Merck) and stored at -20C until further use.

PCR Amplification for an Inhibitory Effect of Ancient DNA Extracts
In order to quantify the inhibitory effect of an ancient DNA extracts plasmid DNA with a known contemporary insert DNA was spiked with diluted ancient DNA extracts. M13 universal primer was used to amplify the plasmid insert in the presence of 1/ 10, 1/100 and 1/200 fold dilutions of ancient DNA extracts.

Amplification and Sequencing of Plasmodium Falciparum DNA
To test for the presence of Plasmodium falciparum DNA in our ancient DNA extracts, PCR primers were designed to target small and partial fragments of apical membrane antigen 1 gene (AMA1) and polymorphic block 2 region of merozoite surface protein 1 gene (MSP1). AMA1 is a protein expressed on the surface of merozoite stage Plasmodium parasite and is highly conserved across all apicomplexa parasites [35,36]. This protein is involved in the merozoite invasion of red blood cells of the host and thus, essential for proliferation and survival of the malarial parasite inside the host. MSP1, another merozoite protein, exhibits extensive antigenic polymorphism among parasite strains/isolates [35,36]. Based on sequence diversity analysis, MSP1 gene is divided into 17 blocks: 7 highly variable blocks are interspersed with five conserved and five semi-conserved regions [36]. The MSP1 allelic variants fall under the 3 major types of allele families termed K1, RO33 and MAD20. Target genes locus were amplified in a 20 ml reaction volume containing 1X PCR buffer (Faststart, Roche), 1.5 mM MgCl2, 200 mM each of dNTPS, 2 mM of each primer (Sigma), 1 Unit of Faststart Taq polymerase (Roche), 2 ml of extracted aDNA and the reaction volume was made up to 20 ml by adding PCR grade water. The cycling conditions for AMAI PCR using GeneAmp ABI 2700 (Applied Biosystems) were: 94uC for 3 min, 45 cycles of: 94uC for 30 sec, 48uC for 30 sec, 72uC for 30 sec and the final extension at 72uC for 5 min. For MSP1 the same cycling condition was used except the annealing temperature, which was at 55uC for 30 sec for all the three allelic families of MSP1 (K1, MAD20 and RO33).
AMA1 and MSP1 gene fragments were amplified in a 20 ml reaction volume containing 1x PCR buffer (Faststart, Roche), 1.5 mM MgCl2, 200 mM each of dNTPS, 2 mM of each primer (Sigma), 1 Unit of Faststart Taq polymerase (Roche), DNA, and the reaction volume was made up to 20 ml by adding PCR grade water. The cycling conditions for AMAI PCR using GeneAmp ABI 2700 (Applied Biosystems) were: 94uC for 3 min, 45 cycles of: 94uC for 30 sec, 48uC for 30 sec, 72uC for 30 sec and the final extension at 72uC for 5 min. For MSP1 the same cycling condition were applied but using 55uC as the annealing temperature for the three MSP1 alleles K1, MAD20 and RO33 ( Table 2).
AMA1 and MSP1 PCR products were cleaned by ExoSAP-IT (USB) and directly used for cycle-sequencing with BigDye terminator v3.1 chemistry (Applied Biosystems). MSP1 PCR products were cloned using TOPO TA cloning kit (Invitrogen) and transformed into E. coli Top10 competent cells. Colonies were PCR-screened and sequenced with M13 universal primers (Eurofins MWG Operon).
Amplification and Sequencing of Mycobacterium tuberculosis DNA PCR amplification was performed using primers P1 and P2, recognizing a 123-bp fragment of the IS6110 sequence of M. tuberculosis (Table 2). This segment is specific for the MTB complex, consisting of Mycobacterium tuberculosis, Mycobacterium bovis and Mycobacterium simiae. IS6110 is present in multiple copies in M. tuberculosis strains and in a single copy in M. bovis strains [37,38]. The nested-PCR method is employed to detect the presence of mycobacterium element. The first pair of primers (P1 and P2) amplifies a 123-bp. The second set of nested primers; (IS-3 and IS-4) partially overlapping P1 and P2 amplifies a 92-bp product ( Table 2). The PCR mixture consisted of 1x PCR buffer (Faststart, Roche), 1 mM MgCl2, 200 mM each of dNTPS, 2 mM of each primer (Sigma), BSA (at a final concentration of 4 mg), 1 Unit of Faststart Taq polymerase (Roche), extracted DNA, and the reaction volume was made up to 10 ml by adding PCR grade water. Amplification steps consisted of an initial denaturation at 94uC for 5 min and 42 cycles of 40 s at 94uC, 1 min at 68uC, and 20 s at 72uC, followed by 5 min at 72uC. For the nested PCR, 1 ml of diluted first-round PCR product was used as the template. Nested amplification was performed using the same thermal profile as before, but applying only 25 PCR cycles in total.
Specificity of the IS6110 PCR was confirmed by restriction digestion and direct sequencing. For restriction digestion, 4 ml of the PCR products were incubated with 10 U SalI (NEB) for 2 hours at 37uC in a water-bath. For direct sequencing, the PCR products were purified using ExoSAP-IT (USB) following the recommended protocol of the manufacturer. Cycle-sequencing was performed with the BigDye terminator sequencing kit (Applied Biosystems). Automatic sequencing was performed on an ABI sequencing model 3100 (Applied Biosystem).